Andreas Klinger | Algora TV

🔥 Investor Open Office Hours - Bring your demos and questions! - Chat @ rfc.to 🎉

Andreas Klinger

Watching now

McPizza hallo hallo

Zafer Cesur 👋

Zafer Cesur *anticipation intensifies*

Richie McIlroy test123

Ioannis Rafail Florokapis India show was fire - excited for today's stream!

Richie McIlroy ayyyy pontus

Ioannis Rafail Florokapis eyyy Pontus let's go!

Andreas Klinger https://midday.ai/pitch

Vlad Hey we're building in the space - to help w roadmapping/exits! Love this! Https://usenomad.ai (we're more on the interest rates and run way side ;))

Vlad How are differentiating from say QB/QBO? Trying to figure out how you're positioning yourself - the target customer?

Richie McIlroy 🔥

Ioannis Rafail Florokapis great points about when OSS is best choice

Vlad Do you intend to build regulation processes / UX on behalf of the Tax office? Huge opportunity there - much of the financial services has poor UX. This is the piece/services we intend to monetize.

Ioannis Rafail Florokapis 'building blocks' infrastructure major key

McPizza amazing story

Ioannis Rafail Florokapis what a founder line-up today 😍

Filipe Alvarenga It's looking awesome, already wanting to try it out!

Andreas Klinger https://uninbox.com/

Filipe Alvarenga niche suggestion: I've worked at a Slack competitor, and their growth came almost entirely from providing a Slack alternative where you can control all your data (it was open source). Their strategy at the time was to federalize their system so it would turn into a protocol and different organizations in different servers would be able to collaborate seamlessly. I think you could pursue something like that. In some sense you're already "federalized" because you run on top of email

Vlad TIL hype gen for b2b

Vlad skiff seems to have done well in the email space - worth for case study

Vlad cybersecurity for emails :)

Filipe Alvarenga sorry for the confusing message, typed in a rush while having lunch xD

Vlad Thank you for the session! Randomly jumped in - great insights!

McPizza Thanks everyone!

Ioannis Rafail Florokapis Personally, if an app took all my inboxes (gmail, company etc) and gave me a nice UX for reviewing categories ex. 'deadlines approaching', 'support', 'accounting', 'newsletters', 'new connections' etc. and was basically an AI secretary + Copilot for my inbox, this would help me a lot

McPizza im always in the chat 🚀

McPizza i cried that night - cried myself to sleep

Andreas Klinger i for one welcome our AI overlord's payments

Ioannis Rafail Florokapis very cool. Agent to Human. Agent to Agent payments

McPizza feels like a missing link for the future problems that we're all gonna face. I love it

Peer Richelsen sup peoples 👋

Peer Richelsen it somehow disappeared on the calendar

Peer Richelsen i think all cancelled

Peer Richelsen lets do it

Peer Richelsen wait for me

Peer Richelsen wait chat is delayed?

McPizza yup

Peer Richelsen i dont know

Andreas Klinger https://tv.algora.io/PeerRich/

Peer Richelsen shit

McPizza 👋

Ioannis Rafail Florokapis 👋

00:00 you you you you you you you you you you you you you you you you you you you Okay. Okay. Okay, let's do this.

06:53 What's happening with my camera?

06:56 Okay, let's do this.

07:01 Okay, works.

07:02 Awesome.

07:10 this. Okay. Works. Awesome. So, um, last week we had a lot of good feedback on the India show.

07:18 Um, and I wanted to say thanks to everybody who joined, uh, especially, um, to my experts and, um, I want to do like a deep content dive, like I don't know one or two weeks let's say every two

07:28 or three weeks but not like every week okay so today what I want to do is like a very simple

07:36 very quick in and out open office hour so the the idea is basically we do just founders, no additional content, no crazy one.

07:48 And yeah, we're trying to have a good time.

07:50 Okay.

07:51 So welcome to request for comments.

07:58 The idea is literally in the name.

08:01 So you guys know the drill.

08:03 Join the comments.

08:04 It's on rfc.tu let me make this a bit

08:08 bigger yeah join the comments it's meant to be highly participatory be nice we still don't have

08:19 a proper way to delete comments so don't want to like encourage anybody but just saying it you know like be nice you guys are like we have done this a few times so you guys know who i am um but i want

08:31 to do like quickly an intro anyway um my name is andreas i have been cto vpe founder of a few

08:39 companies most notably i was founding team and cto at Product Hunt. We got acquired by AngelList where I led engineering teams in the crypto team called

08:50 CoinList, in the talent team, which is like the big jobs board, if you might know.

08:53 And afterwards, there was CTO at an online accelerator called Ondek.

08:59 I have also been running a small micro fund where I need to change the name at some

09:03 point, but I did like roughly 100 investments in the last four years with this fund.

09:09 I think it's 102 now.

09:11 Very global, very early first check, first investor.

09:14 I'm like one of those few people who are very comfortable

09:17 investing without any lead and ideally as the first person.

09:22 Very exciting as an investor.

09:23 I don't know if it's like economically good idea but yeah

09:26 cool so last week we had a lot of good feedback a lot of people

09:31 sent me positive dms on it highly highly appreciate i won't be able to do like deep

09:38 content dives every week i have like multiple full-time jobs i've been coding all day today. So I will try to do have like a bigger

09:46 content deep dive, like every, let's start with two weeks and see if this works and have like a

09:52 bigger topic around those. And in between do like quote unquote normal open office hours where I

09:57 will try to batch founders that have like roughly a common theme. Okay. For the next week i was thinking doing uh something along the lines of

10:07 like ideation finding startups but also like having people pitch their idea for products that

10:13 should exist um so if you think this is a good idea or you want to join like please uh dm me and we'll do it from there. One second.

10:37 Actually, give me one second just to...

10:44 There we go.

10:47 I hope that's the right one um cool so uh today uh as i said before very quickly easy peasy office hour no big content

10:57 blocks no special guests this is still a very much a better run so please let me know what

11:03 you think what could be improved and everything

11:05 i'm mainly doing this as a sort of like exposure therapy because i hate public speaking i hate

11:10 this kind of stuff and i'm constantly afraid that i will like

11:14 like leak my bank details or some shit you know so or like swear um so for me it's fun but like

11:22 also let me know what could be more useful, what we can do

11:25 here, right?

11:26 Cool.

11:27 Rules.

11:28 You have been here, you know this, join the chat, share your POV, ask questions.

11:32 It meant to be very participatory.

11:34 Don't be rude.

11:35 Let's be helpful.

11:36 We're having founders here who are very early sometimes, and I intentionally don't ask for

11:42 pitches.

11:43 Those shouldn't be polished demos.

11:44 Those shouldn't be polished demos, those shouldn't

11:45 be polished speeches and rehearse things. This should be real founders like you, like me,

11:50 like everybody else. This isn't meant to be shark tank. This isn't like some big, weak investor

11:58 telling if you have like something worth your time or not. You know, this is like

12:02 people trying to help each other. Some of them will be quite early we are not doing here we're not like looking for investments we're looking

12:08 for like ways to improve here cool open office hour um let me get the first guy in um

12:19 so did this work?

12:27 Jesus.

12:29 Cool.

12:30 Give me one second.

12:36 I will get you on the main screen.

12:38 So, does this work?

12:42 Screen share?

12:43 Yes, it works.

12:44 Perfect. Hey, mate.

12:45 Hey, man.

12:48 So, hey, hey, hey.

12:51 Can you quickly share your name and what you're working on?

12:55 And we'll take it from there.

12:58 Yeah, so my name is Pontus.

13:00 I'm working at Midday.

13:02 Yeah, I've been doing that for five months while working full time.

13:06 So yeah. Nice. You guys went a little bit viral recently, right?

13:14 Yeah, I think we started with a pitch deck that we made for ourselves basically.

13:21 Instead of doing the regular PDF, we just made this website and yeah basically then it flied

13:26 really high

13:28 we got a lot of bookings

13:30 and yeah now we're here

13:31 can you quickly share the

13:34 pitch deck maybe because I think it's

13:36 like very rarely that like pitch decks get

13:38 viral actually

13:39 yeah so it's

13:42 media.ai

13:43 slash pitch.

13:46 Okay.

13:47 Do you want to screen share?

13:48 That's the easiest, actually.

13:50 Can I do that?

13:51 Let's see.

13:52 Screen share.

13:53 I'm going to see this one.

13:57 Nice.

13:59 This is your actual investor pitch deck, right?

14:02 Yeah, so you can actually see the view counts as you see it.

14:07 Can you make it full screen?

14:08 Quite crazy.

14:09 Let's see.

14:10 Is it me or is everything now in the bottom?

14:18 No, everything is full screen.

14:23 Now then.

14:27 Something broke in your UI.

14:30 Okay.

14:32 Let's see.

14:33 Or make it smaller and just share the browser.

14:36 We have it full screen.

14:38 Yeah, let's do this then.

14:40 Share window, Chrome tab maybe, pitch.

14:44 Yeah, there we have it

14:46 right

14:46 wait am I doing something wrong

14:51 maybe it's on my end one second no wait

14:54 what's happening here fit to screen

14:56 let's try this okay now it works

14:58 okay okay okay

14:59 cool awesome yeah Pontus

15:01 what is midday and like what is the

15:03 deal with a pitch deck that can go viral?

15:06 I mean, Midday is basically a thin layer between your old 90s bank and your accountants

15:12 software, basically to streamline your end-to-end procedures, running your business smaller.

15:17 Yeah, I mean, we used to say SMBs, but we'd rather say micro businesses where you don't

15:23 have anything actually we just

15:25 want to streamline your processes yeah that's what we're aiming for yeah and the the idea so like to

15:34 to quickly explain this to people is basically open source financial software? How would you summarize this?

15:46 I would say that we are in situ for actually how your business is doing financially.

15:52 I will quickly open your website if that's okay.

15:54 Just to let me see what the website is saying.

15:58 So everything open source.

16:00 I see my bank accounts.

16:02 I see where I'm spending money. And I can connect this directly to my bank accounts um i see um we're spending money uh and i can connect this directly

16:07 to my bank accounts and then from here directly to my bookkeeper sorry to my tax advice this is

16:14 correct exactly yeah perfect and i have a virtual cfo yeah i mean what, backtrack a bit, we started building this in public on Twitter,

16:25 and that made us basically take this decision on what we should focus on, basically.

16:32 And from our early users, we really got feedback on, they really want to know how the business

16:38 is doing financially.

16:39 And that led us to like burn rate, runway, and then also, of course, be able to ask questions around your financial situation, basically.

16:49 So, yeah.

16:50 Cool.

16:51 I think that is a...

16:51 Quickly show us the pitch deck because I need my pitch deck to go viral.

16:56 All right.

16:57 So, yeah, this is the start screen.

16:59 You can navigate with your keyboard.

17:01 And the nice thing, actually, which is it's not a PDF.

17:04 If you have spellings where you want to correct things, you can just do that, right?

17:08 Because if it wouldn't be a PDF, you would send it away.

17:10 But yeah, navigate, you can navigate with your mouse also, of course.

17:16 And then you have a actually streamed HLSL streamed video demo and the product.

17:30 And then let's see live stats of our users.

17:33 This is actually streamed in real time from Superbase.

17:36 Just recently hit 100K transactions.

17:37 Nice.

17:41 And then basically just showing what people are saying on Twitter and actually linking out to the top tweets.

17:46 And actually one big reason why it went viral is that Guillermo actually tweeted this not once,

17:52 but three times. So yeah, it went really, really crazy. But yeah, basically everything we want to

18:01 do, clear actions. Victor did this for ourselves.

18:05 What is our purpose?

18:07 What is our mission here?

18:09 Basically, just showcase it.

18:11 Yeah, we took it from there.

18:12 And actually, we had this Cal.com embedding here,

18:16 which made our calendar be fully booked.

18:19 I think that also went quite viral.

18:22 We didn't add any minutes between the meetings.

18:25 So we basically had meetings from back to back for two weeks.

18:30 Yeah, quite crazy.

18:33 And you had like, I'm sharing on my screen quickly, like you had like dozens of signups overnight.

18:39 So like basically TLDR is you have a lot of good problems to report and to wrap in the faces of people here, which I like.

18:49 Yeah.

18:51 So you had like a bunch of calls with investors.

18:53 So two questions.

18:54 What happened and what would be useful to discuss in this call?

18:59 I mean, you and I and Victor actually had a really great quick call before our investment meetings because we didn't know nothing about what it's actually about.

19:11 So we knew that we may want to raise some sort of money, but we realized that we really need to be precise what we actually want to do.

19:25 we realized in the process where we got a lot of great investors that we're going to continue on our end to be fully self-founded to work with our early adopters because we have the luxury here

19:31 that we actually can do that so yeah we we have a lot of great faces a lot of great connections

19:38 with bcs a lot of great feedback actually that we saved for ourselves and we don't close any doors at all.

19:45 We're rather pausing or postpone it a bit.

19:49 And we want to find our velocity, basically.

19:52 As you know, we've been building this while working full time.

19:57 So it makes sense for us to see how far we can take it, just me and Victor, full time.

20:04 What was your um like what is your

20:06 summary or like what it was like how would you summarize the the feedback from the investors

20:10 like how many investors did you meet like in what time period

20:13 uh around 20 to 30 between two weeks

20:19 and the summarize from those meetings basically it's it's they really want us to put us in sort

20:29 of a box some sort of success that already have happened of course like a unicorn or something

20:35 but we realized with all this meaning that we actually want to do a different route because

20:41 we want to be the thin layer between everything we don't we don't want to give you interest rates on bank accounts etc we we want to provide it to for the the smallest

20:51 businesses because we want to grow with them so we learned a lot and we learned that our

20:57 competitors actually who raise money needs and also goes for the more account stuff with interest rates etc and me as a long

21:07 term business owner i just want to stream by my end-to-end processes basically i want to save

21:12 hours and money when you say like they wanted to put you in a box like where like any boxes mentioned

21:19 i mean fintech is it's quite a box, but the usual box is bookkeeping, of course.

21:28 Like, are you going to be a full-fledged bookkeeping system or are you going to be a fintech company where you give interest rates, real cards, etc.?

21:41 And I would say, yeah, we realize we want to be in the middle. So that's basically what

21:47 we concluded. When you say in the middle, you mean more like, before I say it for you,

21:51 like what does it mean? It means like me being in the middle. So yeah, usually when you run your

21:57 business, you already have a bank, right? Or you could potentially have a modern bank like

22:01 Mercury, of course. But in the end, I don't want to change my bank as a business owner.

22:07 I don't have that time.

22:08 I don't have that.

22:09 I can't put my hours to actually change my bank.

22:12 So we want to piggyback on top of your old bank thanks to open banking.

22:16 So you're sucking in all your transactions and you're good to go.

22:20 Basically, the hurdle to get started with me is much, much lower.

22:23 But we can still provide a lot of great tools on top of that.

22:27 So you can keep your old bank.

22:29 You can keep everything that you already have in place, right?

22:32 And then you can start to work with Mide and really start to optimize the inflow,

22:38 invoices that match your transactions.

22:42 That is one aha moment in Lidl

22:45 where our users really say,

22:46 we're saving a lot of hours here.

22:49 So instead of going full-fledged banking solution

22:54 where you need to opt in for a new bank card, etc.,

22:57 we just give you a tool on top of everything you already have.

23:00 What's your take on using open source in this space?

23:04 And what's your take on using open source in this space? And like, what's your take on like, how to say this?

23:08 Like, I feel like when we say B2B, this is like almost like a broken acronym, you know,

23:14 because like a small SMB has nothing to do with like an enterprise use case, you know.

23:18 But like, what's your take on open source here?

23:21 And like advantages for like different kinds of businesses in general what's

23:26 your take on open source for this kind of industry and also like how do you think that these different

23:33 groups of b2b will actually play out in this industry like does it make sense what i'm saying

23:39 yeah i mean there are several reasons why we went open source. Actually, there were no reasons not to go open source

23:45 because we want to build from the bottom up, right?

23:48 And do this as a newcomer.

23:51 Sure, we can build this in stealth mode for six to 12 months

23:55 and say, hey, we're here now.

23:57 But instead, we're flipping the table and working together with our users,

24:00 which means we're going to find product market fit much, much earlier

24:04 and spend a lot of less money on building a stealth mode.

24:10 And extending on that with open source is, of course, trust and transparency.

24:15 And we realized there is no actual financial tools where you build trust in a community

24:21 and building the deeper connection with your customers and your early adopters.

24:28 And I think Twitter has been an eye-opening thing for us to see.

24:32 We can really work with our users and go open source.

24:37 It's basically, yeah, if we can work with great engineers on GitHub, we can really deliver something

24:45 that is truly unique

24:47 like in this area.

24:50 So I don't actually see any real reason

24:54 not to go open source.

24:55 Of course,

24:57 when you ask like,

24:58 aren't you afraid of people

25:00 are gonna steal parts of your code,

25:03 et cetera.

25:03 But I think it's nice to just build something

25:08 and build a brand around it.

25:10 And then in the end,

25:11 if you want to do your own thing on top of it,

25:13 yeah, sure.

25:15 But I think it's much better

25:17 to just be transparent and fully open source.

25:21 Do you quickly want to like five minutes demo

25:23 the current product for one second?

25:25 And I have like two questions related to that. So one from the audience, like from Duda,

25:31 like how is it different to QuickBooks and QBO, like QuickBooks Online? And like,

25:38 so like differently put, like how do you position yourself or like how do you see

25:41 like a target customer segmentation?

25:42 yourself or like how do you see like a target customer segmentation yeah so we are quite different in terms of quick books because we are not the full-fledged book

25:52 keeping software in the end we export where your accountants are sitting and and here in sweden

25:57 you have a quillian name it's fortnox system and those system is basically made for accountants

26:04 and your accountant basically say

26:06 can you please use Fortnox because I need to collect everything to to make your bookings

26:11 and it's the same for the quickbooks exterior etc it's not made for actually running your

26:17 business as seamless as possible those tools are made for actually closing your books together with

26:21 accountants or actually posting those to the tax office so we are a thin layer collecting everything you need and in the end just export

26:30 that to your book software and on top of that you get a lot of more insights based on your financial

26:35 situation so yeah what's your what's your uh take on like so for example like when you said before

26:42 like you talked a lot about trust um my so i'm a big

26:45 believer in all commercial open source and i i see your point on like people potentially stealing

26:50 your source code and so on and so on but like commercial open source is like extremely powerful

26:56 for i would say like almost commoditized industries you know where like every product

27:01 is kinder like you know it's like similarish when it comes to feature set and all this kind of stuff, you know.

27:08 Yeah.

27:09 How do I get you back here?

27:12 But I see the biggest opportunities in the commercial open source usually around self-hosting slash data compliance and potentially also extensibility or just like keeping it alive once the startup is

27:27 like god forbid like no longer operational you know i mean yeah yeah what's your take on this

27:34 like do you think this is like like is this something you're actively working on or planning

27:39 for or like what's your plan here yeah so of course we want to have a one-click self-host solution uh that is our goal of course

27:47 because actually running your your your financial tool on your own servers owning everything you

27:54 know exactly what's happening of course that is a really new way of actually running your business

28:00 so yeah of course that's that's what our ambitions. We're still too early to have that in place, of course.

28:07 We realized that we need to provide some sort of banking providers engine,

28:13 which I mentioned a bit earlier.

28:16 So to be able to actually get the transactions,

28:19 you actually need to get all these contracts from Ladd and Teller

28:23 and GoCardless, et cetera.

28:24 So we actually add on top of that our engine, which means you can use our API instead to

28:30 get your transactions in one unified format, basically.

28:34 So we have all these things in mind to also be extendable in terms of APIs, etc.

28:41 Yeah.

28:42 And in the end, of course, self-hosting is key here.

28:46 Okay. And last question on my end end to get back to the investors so how many investors did you roughly meet like in the last two weeks

28:51 i would say roughly 20 to 30 investors in two weeks

28:57 like it's not like always my main recommendation is if you do fundraising, push everything as close as possible together so that you can like create momentum, you know.

29:08 And so you basically, you A, quickly find out if it's possible or not, you know.

29:15 And B, if it's possible, you have a chance of it becoming like a self-fulfilling prophecy.

29:20 You know what I mean?

29:21 Yeah, yeah, yeah.

29:21 Or if you find out, hey, currently it's like i'm not in a good position

29:25 not in a power position or like like where the market currently is trying to put me doesn't feel

29:29 right you know you can just like yeah very quickly find that out and like move on you know um yeah

29:38 yeah what's your what's your take on like um in general fundraising? So you want to get a little bit closer to product market fit for now?

29:47 Do you have any further thoughts on this?

29:50 Yeah.

29:52 So me and Victor are going full time, as we mentioned today,

29:56 which means we're going to work with this 100% or even more, of course.

30:01 And we have the luxury here where we've been working as a consultant

30:06 for several years.

30:08 So we can continue with this

30:09 without worrying for a while.

30:12 So what we actually said to all the investors

30:15 that shown interest, basically,

30:17 that, hey, we're going to postpone this

30:18 for at least six months

30:20 because we want to work super close

30:22 with our early users.

30:24 And we don't want to to to lose

30:25 momentum doing this um so we're not stressed but we we say that we're not closing any doors right

30:32 now but we're going to continue work for at least six months uh with our setup i mean that's in

30:39 general like uh in in my opinion with uh fundraising uh so i i from a very first startup I raised for like over a year

30:47 you know and if I would have just spent that time instead of like meeting investors just like

30:53 focusing 100% on product you know I would have most likely raised around earlier than a year

30:58 you know I mean yeah so that's like one of my biggest lesson learned you can like investors

31:04 get paid to meet

31:05 you like you can wait you can do this like every day as much as you want you know uh and then next

31:11 month you have like you can do like almost like a cycle and like re-meet the old people you know

31:16 yeah yeah yeah but it's also it was super nice actually meeting all the investors of course

31:21 to get their insights and get their view on things

31:25 because i mean they are they are expert on actually reviewing these kind of things right

31:29 so it was super valuable for us to actually have these meetings and actually see who they are and

31:34 also in in a future where we potentially want to raise money then we know by fact we want to reach

31:41 out to basically um a last question from the audience and then we wrap

31:45 up um do you have any ideas around um wait a minute let me actually like before i read this

31:53 for you this is a longer one but basically the question is around um tax regulation um and

32:00 government infrastructure so basically yeah um do you intend to build regulation processes

32:05 UX on behalf of the tax office?

32:09 Many financial services have like horrible UX

32:12 and this could be like an interesting service

32:14 like a building block, right?

32:17 Yeah, this is actually the key reason

32:20 why we're not going full bookkeeping software

32:22 because then we couldn't go global from day one.

32:24 We couldn't be this innovative on our user interface, etc.

32:29 So we will never be a fully bookkeeping software,

32:32 which means the only things we're actually exporting

32:34 to your accountant or your accountant software

32:37 is transactions and the matched receipt, right?

32:42 And then in the end, you need to have your mark on it

32:44 from an accountant,

32:45 etc. This looks okay. So you're saving the hurdle for actually doing the mandatory test,

32:50 collecting everything, etc. This is like one thing open source, commercial open source especially,

32:57 usually has like a strong opportunity is to think of its own product as like building blocks that people can also like use independently

33:07 you know and almost like use them as infrastructure so for example in your case if people don't like

33:12 your ui for some reason but still want like logic flows around aggregated bank accounts you know

33:18 that would be building blocks that you could offer you know and maybe in future when like you have more time more

33:25 money more everything you offer regulation processes or like tax submissions or like who

33:30 the hell knows what you know um and at that point this could also be an interesting building blocks

33:35 for you yeah i think i think being in the middle layer like being extendable on different angles

33:42 is key here to know because we don't know exactly what

33:45 we're going to do in the long run right we know what we're going to do in the short term

33:48 and our users are going to tell us basically we need this and then we're going to find the

33:52 patterns together with those because we're flipping the table we're not sitting we have

33:55 a public roadmap we have everything you can request features and based on that we're going

34:00 to see the bigger blocks and i think key here is open source and be extendable in terms of apis yes cool awesome uh thanks so much for joining so we have a founder who had a call with

34:13 me and then afterwards realized vc isn't for him um although he had he met like 200 other vcs roughly

34:21 um cool um the keepers in the loop. Is there anything that

34:27 would be useful from the audience for you? Like anything that people can do

34:31 that would be a help for you? I mean post if you find me interesting and

34:36 yeah post what you need and then we're gonna find the patterns together with

34:40 you. Awesome. That's interesting and also personal thank you for both me and

34:44 Victor. You have been really great for us. Likewise, it was awesome to meet you guys. And like one thing that

34:51 I always like to do is like trying to meet people, even if I don't know if I can invest, you know,

34:58 because you're building something really, really interesting, you know, and me i i i think we need to build amazing businesses and make a

35:07 shit ton of money you know and some of them might be good vc cases you know i mean and it's sometimes

35:12 early to say which one is which you know so yeah yeah yeah of course cool thanks so much for your Yeah, thank you. See you around. Bye-bye. Cool. That's double me.

35:31 So, my Zoom is fucking up.

35:35 Isn't this nice?

35:37 Goes completely viral, has several thousand people subscribing,

35:40 several hundred investors wanting to talk to him,

35:43 has one call with me and then afterwards realized

35:44 that VC might not be the right thing that's the kind of magic I do by completely

35:50 alienating people from the investor route let's try you on hey does this work test test hey omer

36:16 test test do you hear me testing testing i hear you loud and clear perfect how's it going very very good um we have a similar case with you right like you

36:30 also went viral recently you had like your product hand launch we had our product launch

36:36 about a week ago um we got number one of the day um and yeah it's just been pretty amazing and pretty epic. Very unexpected as well,

36:47 considering the kind of space that we're in. But yeah, maybe that's a good time to introduce what

36:55 space we're actually in. Screen share and walk us through what you're doing. And also,

37:00 what would be useful for you for this call? What do you want to get out of this?

37:01 And also, what would be useful for you for this call?

37:03 What do you want to get out of this?

37:06 I want to get everything out of it.

37:08 Everything that's possible.

37:12 Let's share the entire screen.

37:12 And there we go.

37:13 Okay.

37:16 So I'm screen sharing now.

37:18 We are on inbox. And on inbox is essentially a modern take on email.

37:23 essentially a modern take on email.

37:28 The challenge that I,

37:30 as a previous startup founder and business owner,

37:31 many,

37:35 many times over email is such a chaotic experience when you're just starting.

37:36 I mean, we just started on inbox and I already have 280 messages and it's chaotic to

37:41 organize chaotic to deal with.

37:44 But also from the point of view of there's no collaboration in native email.

37:49 And the collaboration aspect or the team usage is more of the use case of what we're focusing on.

38:01 So we're coming out with an open source modern alternative for Teams or some people

38:06 say what email should have been from the start. The core challenges that we're trying to address

38:14 are things like collaborating on emails or email conversations with your colleagues or

38:21 your other members of your organization.

38:23 So kind of like Front? or like what does it mean?

38:25 Yeah, very, very similar.

38:27 A lot of very similar use cases to front.

38:29 To clarify on that point, I worked with the team at ClickUp for a year.

38:35 I led their automations and integrations in terms of product, but also led the engineering

38:41 teams before they got bought in engineering managers.

38:44 So I come from a very strong productivity background.

38:49 And Front has a lot of use cases.

38:52 For those who maybe don't know, Front is essentially a shared inbox experience where you can go

38:58 in and you can post comments.

38:58 So we have a lot of the crossover features from there.

39:02 We're taking things fully open source as well.

39:06 So maybe this is a open source stream. And yep, commercial open source. We have our cloud offering. You can sign

39:12 up and join now. But the key things are we're focusing on bringing what we know and expect

39:19 from modern tools today, like Slack, like Figma,ma like whatever the collaborative experience and bringing that into email um so yeah we launched on friday um this is literally the the

39:31 the mmp um like it's just a different email chat experience what is mmp minimum mail program or

39:38 what it means different thing for different people, but generally minimum marketable product,

39:45 i.e. the minimum thing that we can kind of attract people.

39:48 Email is unsexy.

39:50 Email is just that nobody wants to deal with email.

39:53 Okay.

39:54 So we had to find what's the possible way,

39:57 how can we make it as sexy as possible to entice people?

40:00 And this is essentially it.

40:02 We do a lot of work to strip away a lot of the noise from emails

40:05 so I just sent this quickly well as you said to me Omar we're moving your time up like I sent this

40:11 for my google account I have a signature on my google account and the email came through to

40:17 an inbox it arrived in like less than a minute as well so it's super fast and one of the examples

40:23 of what things that we do is we strip away signatures and we have this context panel. We're going for more of a chat look right

40:29 now because we don't have a designer on team and I can't think of any other cool designs.

40:34 But yeah, like we strip away the signature, we strip away additional context, like attachments,

40:39 et cetera, and we move them across to the side here instead of having them in the conversation,

40:43 which then allows you to say, you know, just reply back and experience more of a kind of like

40:47 chat-like experience.

40:50 The challenge that we're having right now is our key focus is on the teams and collaborative

40:57 experience.

40:59 We do already have a lot of the core functionality and features for that.

41:04 We already have groups, so we can have a conversation of the core functionality and features for that. We already have groups,

41:06 so we can have a conversation with user groups. Let's come back here and not this one, not this

41:12 one. One of these guys. Oh, here we go. So this conversation here, for example, is between me

41:17 and the support group. Omar McStaging and the support group. And this person here is John Smith.

41:23 He's a member of the support team and he accessed is John Smith. He's a member of the support team.

41:25 And he accessed to this conversation

41:27 because he's a member of the support team, et cetera.

41:30 So we already have this concept detached of users and groups,

41:33 et cetera, and things.

41:36 We're not a traditional email system.

41:38 We're more in terms of a chat app.

41:41 To get to the point of where we're currently hitting some roadblocks is in terms of the messaging and what our core focus should be initially.

41:55 We had a good successful launch.

41:57 We have, I think, 2,500 or probably 3,000 conversations right now.

42:04 People are having 3,000 different conversations

42:05 in our inbox.

42:06 We have, I think, 1,800 or 1,900 users

42:09 in various organizations.

42:13 Our core focus, though, is on the business use case

42:16 and trying to bring things full circle back around

42:21 to how can we present this in a way

42:24 that highlights the business benefits

42:28 of such a system and which features, like what are we missing in order to present that?

42:38 So yeah, our product hunt brought in a lot of individuals.

42:43 They are great.

42:44 They're making use of a lot of the cool features

42:46 like stripping signatures,

42:47 having a cool chat-like experience for email,

42:51 attachments on the side.

42:52 Those are universal.

42:53 But in terms of a business use case

42:55 and business customers,

42:59 what do we need to...

43:02 How should we...

43:03 Let's rephrase this.

43:04 Sorry, it's a Friday afternoon and i haven't had my pizza

43:07 yet um it's we need to retarget our to our core audience in terms of our marketing material in

43:19 terms of our landing page in terms of a lot of other things in order to ensure that we're

43:23 communicating the right messaging as we're going through and speaking to investors and going through and

43:27 speaking to people we have the core functionality in app we don't have we don't have onboarding

43:34 like our onboarding is literally a link to our doc site and our doc site says welcome to our

43:40 inbox documentation coming soon um i've seen it. Yeah, it's a good joke.

43:45 Should almost put a Rickroll video there,

43:47 but we might get sued.

43:51 We need to find the right,

43:54 we're a solo team.

43:56 It's just myself full time.

43:58 And we have one hire working part time.

44:02 We need to look at the messaging,

44:04 correct some of the mistakes

44:06 maybe or things that we've been saying, clarify the wording in order to a

44:11 attract the right users, but b also when speaking with investors what narrative

44:17 should we use when speaking to them. It's a very different language.

44:20 We're talking about the things at once here. I will

44:24 quickly try to reiterate and you tell me if this is roughly correct.

44:28 So basically, you had a successful launch.

44:31 You had a very good product launch.

44:33 You had a very good, in general, public reception.

44:37 You, in general, are quite good at hype building.

44:39 People are very, very close to everything you do.

44:42 That's picture-perfect built in public, I would say.

44:44 are very very close to everything you do that's like picture perfect built in public i would say your current concern is that a lot of the use case of the people that you're currently having

44:50 is more like single player and you're thinking that most likely the or like a recent thinking

44:57 of yours is that most likely the bigger money is like when you have multiplayer for team use cases

45:01 and that could be a little bit more like an interesting infrastructure kind of like an interesting solution for smaller teams that's like question

45:11 number one basically around this concept like around this insight like the question is like

45:15 how do you transition this and so on and so on you know and the second question is what is a good

45:21 framing for all of this for investors because, what is the right angle to these investors,

45:27 like for investors, basically.

45:29 Make sense?

45:30 Correct.

45:30 Okay, cool.

45:31 Yeah, for the second part,

45:32 like the feedback that I'm finding

45:34 is that email isn't sexy.

45:36 Oh yeah, 100%.

45:37 In today's day and age with AI and robotics

45:41 and everything else, email is not sexy.

45:44 But it's a big industry that

45:45 has a lot of potential in it yes um so yeah finding the right messaging to speak to investors about

45:49 this um so let's actually start with this one right uh with the with the uh investors and

45:56 everything i will quickly share my screen um actually how do i do this now one second this way

46:02 and now let me quickly share my screen.

46:07 I will kick you out of screen sharing.

46:08 I'm very, very sorry.

46:09 Sure.

46:15 So the main problem from my point of view is not that it's unsexy.

46:16 Unsexy is cool. Like if unsexy makes a lot of money and you're like the only person who is legally allowed to offer this service, you know, nobody has a problem with unsexy.

46:24 You know, the problem that

46:26 you have is it's too similar to a lot of stuff that exists so nobody believes like that with

46:32 your current angle and plus the one little thing that's different you know it would be like a huge

46:37 differentiation you know i mean and that's more likely the problem. One second. How do I...

46:46 Page.

46:47 Here we go.

46:48 And we discussed this in a recent stream as well.

46:55 But let me quickly reiterate this for people who are first time here.

47:01 If you think of every deal that an investor sees as a 3D space,

47:05 you have execution, like working your ass off,

47:08 like how far did you already get, you know?

47:11 Innovation, like is this something that exists or is like different

47:14 or like fundamentally new?

47:15 And then you have credentials, okay?

47:17 The typical problem that you have is very frequently when people build something,

47:23 it's very much like on the low end here.

47:26 Like it's the innovation is low and you can like work as much as you want.

47:29 But like still the innovation is like it's very similar to stuff that exists, you know.

47:36 The problem you have here is that most investors are not looking necessarily for something that they're looking for outliers and to be an outlier

47:47 ideally it's something that people haven't tried before i think that's the easiest summary right

47:53 and if it's something that has been done like eight times it's very comparable to stuff that

47:59 exists and like being comparable is always hard in VC. Like the moment you're comparable,

48:06 the moment you're comparable, people have analogs.

48:09 They can look like, okay, there is, I don't know,

48:12 RavenMail in India.

48:13 There is like 500 other mails in everywhere

48:16 that are maybe open source, that are maybe T-mails, whatever.

48:20 Like how big did they get, right?

48:24 And if there isn't like a big new why now what changed

48:28 new technology new moment new market shift new technology consumer technology new interfaces

48:35 whatever that you can jump into you know it will be roughly like very much here on the low end you know so what's happening is you need to execute a

48:47 ton you know and execution is usually the one thing that most investors really struggle evaluating

48:54 especially yeah it's just hard you know uh because like what is like good growth rate early on what

49:00 is good progress early on is like 1 1000 waitlist customer paying customer like does

49:07 this mean anything you know it's just hard to evaluate for a lot of early stage and it's also

49:11 hard for you to do this you know so um when we talk about fundraising in general my pov on this this is always fundraising equals narrative x momentum okay so or let's switch to win or

49:30 differently put uh you already had like very very good momentum in your round you know like in your

49:36 current this moment like you had like a lot of bus going on and so on and so on and maybe you had

49:41 like two weeks where you talk to a lot of investors you know the main

49:45 problem is like what is a narrative that you can make um work for you okay and that's something

49:51 when we talk about raising investment that's something that's in your case especially important

49:56 because it's like very very close to stuff that exists yeah make sense so far yeah definitely um

50:01 Yeah. Make sense so far?

50:03 Yeah, definitely.

50:05 My personal POV on this is don't race yet.

50:11 Use the current time to properly explore the product and see if you find something that really, really clicks.

50:19 That's my personal take here.

50:22 We can talk what could be possible narratives.

50:27 So one in your case, let's say this is your whole product.

50:32 Email is like something that every company, every software needs.

50:37 And if you think of yourself as like multiple components,

50:41 and then on top is the UI,

50:44 maybe there's an argumentation that some of that is like interesting infrastructure. components and then on top is the UI.

50:45 Maybe there's an argumentation that some of that is like interesting infrastructure.

50:50 You know what I mean?

50:53 Infrastructure as in we provide it as...

50:56 As building blocks basically.

50:57 As building blocks, basically API style system. The thinking here is I might not believe that you can build another email service that will

51:12 be like 100 times bigger than Outlook and others.

51:19 So the UI thing I might not believe because it will be one of many UIs, but maybe there

51:25 is something in the individual building blocks that makes sense to centralize streamlining

51:32 companies and so on and so on.

51:34 And here we're talking like obvious ones could be if every company needs email, if every

51:40 company needs email, you know, what are the different emails they have you know

51:47 and we're talking here you know like app uh select transactional you know yeah uh we're talking

51:54 customer care we're talking newsletter and so on and so on like there's like multiple emails that

52:00 uh companies have and there might be an interesting argumentation to say like

52:05 transactional newsletter, customer care, and all touch points with the customer should

52:10 go through one system. So that you have like one paper trail, you have like one way to

52:14 see what happened and all this kind of stuff, you know? So maybe there is an angle there.

52:20 I don't know. That's something, I mean, by the way, I'm just like brainstorming here.

52:23 I have no idea, you know?

52:24 That's very. I mean, by the way, I'm just like brainstorming here. I have no idea. You know? It was very interesting. Yeah. So that could be something, you know, that you have like

52:29 one API, one central tool, one log file and so on and so on. That could be potentially narrative.

52:36 Now the question is like, is that big enough? I don't know. Okay. The problem is VCs are very

52:42 flaky. VCs are running after trends.

52:45 VCs don't know what they like until they see it in a lot of cases.

52:49 So my personal take is usually to not focus yet on the VC case,

52:56 but focus first on the go-to-market.

53:00 The way how I always think about this,

53:02 there's like an actual go-to-market in the sense of like, what are we doing right now?

53:07 You know, then there is like an, let's say like a vision, you know, or like short term vision, you know, like what is the product scope that we are actually trying to build right now?

53:19 You know, then there is like my vision, which is like the, if I have infinite time, what else could it be?

53:27 You know?

53:28 And then there is like the bullshit VC vision, you know?

53:33 And those can be like bigger and bigger and bigger.

53:37 You know what I mean?

53:38 Okay.

53:39 And my personal gut feeling is like right now, I would completely focus only on these two.

53:46 As long as you can.

53:47 That is my take.

53:49 What would you think?

53:51 I agree.

53:53 So the longer term vision, like we have just this like la la idea, like the 10 year plan, whatever it may be.

54:01 I think that kind of covers like MyVision and the bullshit VC vision

54:06 for those who want to imagine.

54:08 There's a lot of good stuff

54:09 in here. I think

54:11 what happened was during our initial

54:13 release in a rush to

54:15 not in a rush, but in

54:17 order to kind of appeal to the

54:20 majority of possible people,

54:22 we specifically kind of highlighted

54:24 the features that would be

54:26 interesting to more interesting to individuals rather than it being more interesting for the

54:33 collaborative use cases or for a business yes makes sense i mean i don't think this

54:38 is necessarily a bad thing you know the problem that you have is um

54:43 um and i will read this comment in a second we got like

54:47 a really long comment um the problem that that you have or differently put um it's not necessarily

54:54 like a bad thing to have any kind of users you know yeah like it's a super early product everything

55:00 is broken everything is early nothing really works like let's be grateful for anybody we can get, okay?

55:06 So the challenge for you rather is if that's the big B2B use case,

55:13 you know, where you have like multiple concerns of like, you know,

55:20 stability to start, security, holy shit um ux ui like uh and 500 other things you know like

55:30 companies will have like a lot of big use cases so if you ask me to do my customer care

55:36 through you you know versus like another tool that i can buy for like 20 a month you know yeah

55:43 i need to have a lot of faith that you have

55:45 like tons of things covered, you know, and if for you, it's just hard to build towards this whole

55:52 use case. Right. So what might be interesting is if you currently have like some customer,

56:00 maybe there is like use cases are similar enough, you know, like something that

56:06 people like, what are little use cases that you can do now, even if they don't want to

56:12 switch their main email to you, that they can use you for all the time.

56:16 You know what I mean?

56:18 Okay.

56:19 Yep.

56:20 So basically what I'm trying to say is like you want to, it's always sexist and engineered

56:27 to build like the huge thing, you know.

56:31 What you actually want is you want to pick like a big part of that that's important,

56:35 like a core part and find a way that like you can have it earlier already, you know.

56:39 And the problem is email is so fundamental to a lot of people that you might need to even start here with like an

56:46 niche use case almost you know or like a few of those like niche use cases as the default

56:53 thing people use and a few crazy people use like the full thing here already you know

56:59 yeah that's that i think that's that's kind of what we were planning, or that's what we did for the beginning,

57:07 is just kind of go for the really niche or very hard personal pain points

57:10 and bring people in there.

57:13 And then, obviously, things like will try out.

57:15 So we have 1,800, 1,900 users,

57:18 but that's only like 30 custom domains, as an example.

57:22 Like everyone's using our free domain

57:23 that they can test things. But yeah, I like the idea of our free domain that they can use to test things but

57:25 yeah i like i like the idea of of choosing maybe a couple of specific use cases and expanding

57:31 on that and bringing up some of the messaging for that and finally writing the documentation

57:36 as well for those um to highlight uh those specific it always like depends on your on your

57:44 shipping velocity, right?

57:45 If you think of your whole roadmap

57:47 and building email is one of those

57:50 grind tools where the baseline

57:52 expectation is extremely high, especially

57:54 around professional use case

57:56 email, right?

57:57 And if you break this down at how

57:59 quickly can you actually ship parts

58:02 of that,

58:03 the realistic thing that you want

58:06 to get to is like you want to have um little bit smaller chunks that people can actually already

58:11 start using and i don't know what that could be in your case but maybe it's like something like

58:16 email forwards you know or like um like not throw away emails but like an email like can you like

58:24 you make it very easy for me to have

58:26 like virtual email addresses or something email aliasing etc yeah yeah yeah we've yeah

58:33 there's that's really good i'm gonna i'm gonna take that and and eat some pizza on it um and

58:42 come up with the right.

58:47 Yeah, it's very valuable.

58:51 I think we did run into the situation where we lost focus on some of the key things and it became a big mess of here's a giant backlog of 5,000 possible things that we can do

58:58 and no clear way to prioritize or organize that backlog.

59:05 I mean, that's the hard one in your case, right?

59:07 Like you have, to some extent, you have, you want to get here, right?

59:14 Yeah.

59:15 And you need to build features.

59:17 Like you're basically building features for like the, how did we call it before?

59:22 Like short-term vision or product scope you know yeah

59:25 um let's actually up here like you you want to like build features for this here you know

59:31 but realistically the first people will only use like a small fraction so you're like this as well

59:37 and then to some extent you also want to like already start building for the b2b

59:42 so the hard challenge that you have is kind of like a coming up with like good use cases that i can convince a few freaks to use your thing

59:50 daily or like regularly how many freaks is a few freaks for you let me check how many we have in

59:56 chat how many freaks are in chat so we have like a few hundred in watching right now and like uh

01:00:03 in this in the chat are like 15 people,

01:00:06 you know.

01:00:07 So like how can we convince 15 freaks to actually use your tool regularly for something, but

01:00:13 like take that at least serious, you know, like email aliases and this kind of stuff.

01:00:17 And then how can you get like a subsection of that to use your full consumer scope, you

01:00:23 know, while you're building that out um and build that

01:00:28 out always like with this in mind so like you build consumer like single player you always

01:00:32 build with the functionality of having other people joining discussions you know um so maybe

01:00:38 there is an interesting angle here to rethink uh communicating around emails even if people are not in the same

01:00:46 company you know so i can pull you into a discussion and show you the full discussion

01:00:51 you know that i had with somebody else even though you're not work with me normally you know i mean

01:00:57 maybe there are other use cases like this but like what can you do to get those people to regularly use you earlier we had also like a few other suggestions um uh felipe um also suggested like he worked

01:01:10 for a slack competitor that was open source and um their main okay new suggestion i've

01:01:20 worked the slack competitor the growth came almost entirely from providing um

01:01:23 I've worked with a Slack competitor and their growth came almost entirely from providing

01:01:30 a Slack alternative where you could control all your data. It was open source. Their strategy at the time was to federalize their system so it would turn into a protocol

01:01:35 and different organizations and different servers would be able to collaborate seamlessly. I think

01:01:39 you could pursue something like that in a sense you're already federalized because it's on top

01:01:43 of email. So basically the idea here is is i think there's like multiple ideas here like one is like owning

01:01:50 the data and the second one is having organizations collaborate you know uh which is almost like even

01:01:56 bigger than this one it's like kind of like here you know it's like yeah that's our that's our like

01:02:02 five five year thing of basically building our own,

01:02:07 essentially our own network, our own standards and everything,

01:02:10 and having the cross-communication between everything.

01:02:13 Like a dedicated protocol to replace email

01:02:16 that can work backwards compatible with email, in a sense.

01:02:19 That's like our cuckoo vision, as I like to call it.

01:02:24 When I have enough pizza, enough time and enough money.

01:02:27 I always think like this, absurdly big visions are really, really, really good to help you looking at your product from a different angle.

01:02:37 You know what I mean?

01:02:38 Because if you're always like focus on this here and focus on this here, you know, like you're thinking like this all of a sudden like you're really narrow-minded you know yeah and sometimes if you're like hey why is enterprise

01:02:50 email actually not public you know wouldn't it be way easier to have customer care if everybody

01:02:56 could see that not only the team or something like that like some crazy idea and those crazy

01:03:00 ideas sometimes lead somewhere like i always believe that you need to do two stupid ideas to have one good

01:03:05 idea.

01:03:05 You know what I mean?

01:03:06 So like this,

01:03:07 this enormously big visions are like very,

01:03:10 very useful for this.

01:03:11 I think the controlling of the data part here is like the most interesting

01:03:15 aspect,

01:03:16 you know,

01:03:16 and that also could go kind of like here.

01:03:20 Maybe it makes sense to have like,

01:03:23 maybe your,

01:03:24 your approach is not like from the ui point or like use

01:03:27 cases like this but maybe your approach is more like what if uh my my first enterprise consumers

01:03:33 only use this part in the button that's around like every email that like goes through any system

01:03:40 that you use touches this software of ours and you have a full paper trail on this you know yeah

01:03:46 yeah like an audit an audit log or a vault or something yeah um yeah i mean that's there's a lot

01:03:53 there's a lot to think about but i um i agree that kind of focusing on on the smaller use cases now

01:04:01 while maintaining the vision in the backend to help drive some of

01:04:06 the decisions of what we build

01:04:08 how we build. We don't need to build

01:04:09 for the future vision right now but it's good

01:04:12 to be aware of it because it may change some

01:04:14 decisions that we make now that will open up

01:04:16 the doors later on.

01:04:18 But yeah, it's

01:04:19 super valuable and super interesting.

01:04:29 We have a lot of work to do yeah also like by the way thanks for sharing your current state like i always know it's like kind of uncomfortable to

01:04:33 show like the product that's not yet there where you want it to be you know i mean uh but you know

01:04:39 i mean at the same time um like this is open source. It's fully self-hostable.

01:04:46 I mean, we're going with commercial open source.

01:04:48 So there's some modules that will be locked away behind a license,

01:04:52 things related to billing or cross whatever other systems.

01:04:56 But the core functionality is fully open source.

01:04:58 And at the same time, yeah, this is super early.

01:05:03 It's nowhere near what i would be happy sharing and

01:05:07 exposing but at the same time that's the nature of open source yep makes sense i will kick you

01:05:14 out so we have a streamlined session today uh thanks so much for joining dude thank you very

01:05:18 much see you later cool like omer is like one of these people who are really, really good in building in public.

01:05:29 People underestimate how valuable this is. He has like a fan base already now that's helping

01:05:36 him to launch on Product Hunt and everything else, you know. So being that transparent,

01:05:43 being like that close with your audience and your consumer base

01:05:46 is like extremely powerful um let's get quickly the next person on till

01:05:58 hey boss

01:06:03 do you hear me?

01:06:05 Wait a minute.

01:06:08 Yeah, he hears me.

01:06:11 Okay.

01:06:12 He hears me.

01:06:13 How are you doing?

01:06:14 Hey, I'm doing good.

01:06:15 This was really cool to watch.

01:06:16 I liked the previous demo, too.

01:06:17 The homework is awesome.

01:06:18 Yeah.

01:06:19 That was sweet.

01:06:20 Last week, we actually last week week he was actually in the chat uh this is

01:06:28 like uh and we had another guy having an um ai tool to build apps and as a demo that guy just

01:06:37 like okay you know what let's just like copy omar's app in 20 minutes and see if it works that was like a little bit terrifying yeah yeah no i

01:06:49 love it um tell me about payment like what are you building um maybe you can like screen share

01:06:54 or screen demo or something like this and what would be useful for you um i guess i could just

01:07:00 share yeah people what i've built so far maybe get some thoughts or things like that.

01:07:22 So let me get this going. Okay. Cool. So what Payment does is Payment enables AI agents to pay humans for tasks that agents can't do.

01:07:28 So, yeah, yeah, that's exactly what it is.

01:07:35 In the future, hopefully there will always be things that as AIs and humans work together will rely on each other.

01:07:38 And there might be always things that we do better than them.

01:07:44 If not, I say just start building your bunker today and we're you know get ready for that future but yeah uh

01:07:46 the thought process is uh everyone thinks that we're just going to be paying agents to do work

01:07:52 for us this kind of flips it on its head where if agents keep getting better and smarter um and more

01:07:59 capable why not give agents the ability to pay humans to do tasks? Okay, so just to repeat this, okay?

01:08:06 So you don't know if we will have AGI or not, but at least if we have AGI, you want to make sure that we get minimum payment.

01:08:15 Like minimum hours. This is the idea, right?

01:08:18 Exactly. And when you think of it, if we think of incentives even in that sense like the one of the best

01:08:25 incentivizers is capital well like give the agents ability to spend capital and then like let's

01:08:32 incentivize you know that'll hopefully they still view humans as uh valuable and i think they will

01:08:38 even the smartest people today like what we would consider you could create an AGI into this super smart human,

01:08:46 they still have to work with people.

01:08:48 If they're like delegators of tasks, they still pay people.

01:08:52 So we see that today.

01:08:53 I don't think that it'll change over time.

01:08:56 And I think I want to see humans win too.

01:08:59 So this is where like payment was brought up.

01:09:04 So how does it work yeah so there's three

01:09:08 things that make this possible the first thing when you really start thinking about it is okay

01:09:13 i want an agent to pay a human to do something well how do i like go about doing that well you

01:09:18 quickly get into the thing like how does an agent even set up a bank account like can i can an agent

01:09:24 just go into chase and be

01:09:26 like on the website and be like let me set this up or on mercury or something well no not really

01:09:30 there's like regulatory it's not a human like there's no kyb no k you know all these other

01:09:35 things okay kkyb know your bot but yeah i know your bot kyai one of the other things so you can't

01:09:42 do that easily so it's like okay well let's give them the ability to do that easily.

01:09:47 Yeah.

01:09:47 So the first thing Payment does is it allows you to add your agent here.

01:09:54 So let's say I want to add, I have, let's think of a different one.

01:10:00 I have a developer.

01:10:01 I have a product.

01:10:02 I have an influencer agent.

01:10:04 Let's have like a film

01:10:05 agent or a movie creator agent and the agent description is creates full-on feature films

01:10:15 okay so we now have this agent and we have this agent put into payment. This would be an agent that

01:10:25 someone's already built. They would just add this to payment so that they could actually start using

01:10:31 the API calls from payment with this agent. So this kind of acts as like a de facto wallet

01:10:38 for the agent. And in like the movie creator agent, we can think of the example of, let's say you have an agent

01:10:46 that actually produces in the future, Sora on steroids, where it produces feature films.

01:10:53 Well, what if you don't want the agent to write the script? What if you want the script to be

01:10:59 outsourced to a human and have them compete to basically be like, we'll write the script,

01:11:03 but the agent will actually develop the film. this case you would plug in something you you would use payment to basically

01:11:11 add this tool to the agent which is if you're familiar with like agents and tools tools really

01:11:17 like it's to do and to like in the sense of like langchain right correct yeah like length this this

01:11:23 right now is langchain tools any

01:11:26 agent can have any kind of custom tool and payment is in my eyes the most powerful tool an agent can

01:11:32 have because it gives them the ability to actually not just deploy capital but use capital for labor

01:11:36 so they you would go to payment you would copy this code and this would essentially paste this

01:11:43 into you would paste this into if you're using a link chain agent right now, uh, it would know to call this when it's

01:11:49 been triggered with like, Hey, you have a budget of $10,000. Go find a script from a human that

01:11:55 we can like, based on these categories. Like I want this movie to be about James Bond. I want

01:12:00 it to feature Idris Elba X, Y,Z. And then they can write a script for me.

01:12:05 And then what would happen is you would post the task.

01:12:09 But before you can post the task, you need to add funds.

01:12:12 So what payment allows you to do is, hey, let's just add some funds right now real quick.

01:12:18 So it's funny.

01:12:20 I have a crypto background and a lot of people give me flack for not doing this in crypto right now.

01:12:25 And I think there's a lot of reasons why not to do it in crypto currently.

01:12:30 I looked it up, you were like death row for like a bunch of like very famous projects.

01:12:35 Yeah, yeah.

01:12:35 So it's, I think that my mindset is start with fiat and then over time switch to crypto.

01:12:42 When we look at like when most people get into crypto, it's usually when people can

01:12:48 make money.

01:12:52 This tool gives people the ability to make money.

01:12:54 So streamlining into crypto is really easy over time.

01:12:56 So maybe like crypto will be more useful for like agent to agent payments or something

01:13:00 like this, you know?

01:13:01 So let's see.

01:13:02 Yeah, I think that's a perfect use case for crypto okay so

01:13:07 now what i'll do is i'll pay five hundred dollars and what this is doing is it's saying okay we're

01:13:14 going to allocate five hundred dollars now to this movie creator agent but i think if i refresh it refresh. It should be there. Cool. So now this agent has $500 to spend. And what I'll

01:13:27 do now is let me see. So let's actually start to create the task. I'll do this on my end

01:13:36 over here on my code base. But let me know if you have questions while I'm doing this.

01:13:44 So yeah, my main question is how do you find the humans? Do you have farms know if you have questions while I'm doing this. My main question is, how do you find the humans?

01:13:47 Do you have farms?

01:13:48 Do you have kidnapping?

01:13:51 What's your plan here?

01:13:54 Is it metric style?

01:13:56 Am I in a little waterbed sleeping and doing tasks?

01:13:59 What's your setup?

01:14:01 I think that the tasks will be different based on what the agents,

01:14:07 like I think it's going to feed in both ways.

01:14:09 I think agents are going to see

01:14:10 what kind of humans are available.

01:14:13 And then people building agents

01:14:16 will build agents based on the workforce

01:14:17 available for them.

01:14:19 And then on the flip side,

01:14:21 agents will look at humans and be like,

01:14:24 okay, I'll use one

01:14:26 example of someone that reached out to me, like the largest, one of the largest workforces

01:14:31 of mechanics reaches out.

01:14:34 We think about it today, like there's not really any agents, or at least that I know

01:14:38 of many agents that are focused on like car repairs, you know, because you think how can

01:14:43 an agent repair a car?

01:14:44 They can't do that.

01:14:45 They don't have, unless it's a robot, they'll, they'll go or if they control all this.

01:14:50 Well, if you think of giving an agent capital, you're giving an agent a body,

01:14:53 then if they can spend it. So then you could think of doing is like, oh, I go to payment,

01:14:57 I see there are all these mechanics that are available in the workforce. I'm going to build

01:15:01 an agent that's focused on the automotive side of things. So

01:15:05 let's say in the future, this is a free app for anyone who wants to build it. You can build an

01:15:09 app that basically uses like a chat GPT type approach. Hey, there's something wrong with my

01:15:13 car. You take a picture of it. The AI rent, your agent renders what's wrong with the car.

01:15:19 Use the agent ask, can you give me a budget of $5,000? I can go and fix this. Then you say,

01:15:24 boom, you fill it up on payment. The agent gets agent gets funds then that that agent hires out a mechanic

01:15:29 for you to actually come back and fix the car okay makes sense so um a part of the demo like

01:15:36 what would be useful for you for this call like what kind of feedback or like what kind of challenge do you currently have? I think what would be good to know is what kind of, if there are any agent developers

01:15:52 or anyone here, like what are some like initial use cases that they might want to use this?

01:15:59 You know, I don't know if the chat is active on like with that front, but that would be

01:16:03 really great to know i um so i built

01:16:07 like sometimes like smaller agents with like crew ai you know um and uh like crew.com um

01:16:14 one of the things that could be interesting is if you have a way so like if i understood

01:16:20 currently like your current idea is to have like all the humans in your system, right? Both. I think we kind of touched on this a little bit, Andre. It was like you have an expert network

01:16:34 that you can go filter out to like existing marketplaces and then Payment would also have

01:16:40 its own tailor-made marketplace. So it would operate at work like as a router of tasks

01:16:46 perfect too yeah so so i think like just having a service of this routing itself is already valuable

01:16:55 and then being able to place tasks there you know that's clearly valuable so like but not so much

01:17:02 like in the sense of like the humans that you have, but like me being able to say, um, I have an agent in my crew that, uh, has evaluated that

01:17:12 a certain task is like better, maybe like, it makes sense for you to have like an, some sort

01:17:17 of like tool to decide if a task would be better suited for a human. And yes for which kind of human you know and the ability to connect

01:17:25 to this human yeah that that fits very very well in the crew concept for example like you would

01:17:32 have basically one person who would be like a foreman or something like this you know and be

01:17:37 like okay this is agent work this is human work you know i what I mean? Yeah. So evaluating if something is better done by human, and then evaluating

01:17:49 where this should be routed to, this would be two very, very good API

01:17:53 endpoints. Or like agents, I think you can wrap them as full tools actually.

01:17:57 That's a great point.

01:18:00 I think the other piece that I've spent a lot of time on is like the verification of work

01:18:06 yes and just think through like how do you verify that i've got ideas around that um but any any

01:18:15 thoughts there my gut feeling here is like i think it's already useful if you are so by default

01:18:22 at least for the next foreseeable future if i have anything that's

01:18:26 done by agents even if i pipe it out to humans i don't expect perfection you know if it like if i

01:18:32 expect perfection i would do it myself kind of you know so like there's a little bit of like a fault

01:18:36 tolerance here i think it would be useful in the tool if you can pass directly some sort of like

01:18:41 verification prompt as well but like you have like by default

01:18:46 one like you have by default like a very very good verification process going on you know

01:18:50 to filter like the 80 percent of the wrong cases you know but then I can additionally maybe

01:18:58 you give me the sense of shared ownership you give me the sense of control by like having a little prompt that i can pass on as like a verification prompt you know um like what would make a good submission

01:19:11 you know what would make a good result and then i can like provide you something i think that could

01:19:16 be like a very very smooth way to kind of sidestep that problem almost does it make sense yes i think

01:19:23 um god i just had something there.

01:19:30 Can't remember. I feel like I have like both humans and agents in my minds all the time.

01:19:38 Yeah. Yeah. That makes sense. The other angle here that's really, like, I think we talked about

01:19:44 this, like in our first call, the other angle that could be really, like, I think we talked about this, like, in our first call.

01:19:46 The other angle that could be really interesting here is, like, you basically have two to three products here, right?

01:19:51 Like, you have this whole work, dispatching, routing.

01:19:53 Like, maybe it's four.

01:19:54 Like, you have evaluating if work is human needed, you know.

01:19:59 You have dispatching and routing, you know, of the actual work.

01:20:02 And then, like, verification and everything.

01:20:04 That's, like, one queue routing you know of the actual work and then like verification and everything that's like one queue you know then potentially there might be something where you just like have

01:20:10 the wallet as like its own thing right where you can basically say um give me a virtual credit card

01:20:17 for this um like some random payment i have to do but i still route it through payment you know i

01:20:24 mean even though if the

01:20:25 work isn't done by payment with the idea that I won't want don't want to have like eight different

01:20:30 services that control my wallets for my bots you know I mean yes so like something here where the

01:20:39 agent can say like some agent that I built can say I need a virtual credit card to pay something or like I need some way to maybe that's like maybe I assume personally I assume that Stripe in let's say the

01:20:50 next five months will offer like a way for bots to pay you know I mean pretty certain about that

01:20:58 and that would be interesting if then I can use your thing as like my default wallet for, you know, different things like throw money in through my career, throw money in for crypto, throw money in for Stripe.

01:21:12 Maybe I have actual credit cards in there.

01:21:14 Doesn't matter, you know.

01:21:16 But I have like one service where I can say these agents have like maximum 500 and they cannot use it for weapons and drugs.

01:21:24 You know what I mean?

01:21:24 Because you never know. I know. I've thought of that. have like maximum 500 and they cannot use it for weapons and drugs you know i mean because you

01:21:25 never know i know i've thought of that i'm like the anytime you get money involved for paying

01:21:31 someone it's like okay they could think of something so it's like we also have to be the

01:21:35 filter as well for what they're spending their things an agent wallet is a super cool idea. And that's like where I think this evolves into.

01:21:45 Especially over time.

01:21:48 I can't think of a more valuable use case than agents paying humans.

01:21:54 Maybe they can pay other.

01:21:56 Or pay services in general.

01:21:57 The problem you have is this is like a fast moving space.

01:22:03 And yes, you will have by default value but just like

01:22:06 wrapping any kind of like payment and like having one gatement and validations and like whatever

01:22:11 you know no matter what the payment is for what it's used for but it will be a space that will

01:22:16 be very noisy and like all the big players will try to go in there as well you know so i wouldn't

01:22:21 do this as the only thing i think the dispatching and routing of human work is also really interesting.

01:22:25 And, um, the, the third slash fourth service that could be interesting here.

01:22:29 That, um, that's also obvious is like bot to human communication.

01:22:33 So, uh, a human has like, Hey, I should do this task.

01:22:37 I don't really get it.

01:22:38 Here's a question I have, and that gets easily routed to my bots.

01:22:41 You know what I mean?

01:22:42 Um, and then I maybe have like also uh log files where

01:22:46 i can exactly see what that kind of communication was you know um i think interaction like oh this

01:22:54 is what i was going to mention earlier i think um on that front like the communication with the bot

01:23:00 i think we're seeing more and more like agent to agent conversations

01:23:30 I think we're seeing more and more like agent to agent conversations. I think that over time, as this expands, as like a tool like payment becomes something that, sorry, as it becomes something that like starts looking at work, and it starts looking at like, should I exchange these funds that this agent asked for for this work. I think what we'll start seeing is payments agents will, and this is where I think like ZK proofs can like kind of come into play with agents where it's like, it's talking back to the agent that submitted the task. And it's basically

01:23:35 checking the work that exists. And it's like, Hey, did you want like this number of detail?

01:23:41 It's like, yes or no. And then that submission will say, yes, it like, you know detail it's like yes or no and then that submission will say yes it like you know it'll just show the the green like yes it does that way the agent never knows what the work

01:23:49 is but it knows that it's all the standards of what it's looking for and then that will be sent

01:23:54 back and they're just talking to each other back and forth yes that's interesting actually i i

01:23:59 think you can uh okay yeah that makes sense yeah like worst case you can even um have like an agent endpoint

01:24:09 similar to like the communication endpoint you have like an agent endpoint where they can just

01:24:12 like verify the result and you can just assume that if i do tasks on scale like you anyway only

01:24:19 make money with bots that run on scale and it's anyway my own bots very likely to be you know i mean uh

01:24:26 so i pay you to have my wallets and like support my bots you know so i can assume that my bots

01:24:33 won't try to trick you know i mean like they won't say regularly this is like

01:24:37 bad work although it's they use it you know because otherwise you can just like kick that

01:24:44 person off the platform realistically

01:24:46 you know i mean um and if i do this on scale that's the whole point of it you know it's like

01:24:50 very likely that that's not worth the effort so we're almost like in an area where you cannot

01:24:55 trust the the payers you know but it's at least something where you can have um where you can

01:25:02 assume that people are a little bit more like lazy and good willing you know i mean so if you have like a verification endpoint and be like hey this is the

01:25:08 first part of the submission you know is this within the quality benchmark yes if yes then i

01:25:15 send you the rest maybe you send all of it you know but that's also like really really cool kind

01:25:19 of like the webhooks around um verification if work is what people want.

01:25:26 And this loop, you have

01:25:28 the internal ones, then you have the prompt that they

01:25:30 provide, and then you have maybe also

01:25:32 optional sanity checks

01:25:33 before the work is actually accepted.

01:25:36 Yeah.

01:25:38 And I think you could get various, even on the

01:25:40 Verifier part, you can get specific as

01:25:42 certain, if it's

01:25:44 a developer agent that wants

01:25:45 coding reviews it's just trained on just that like i i don't know i think that agents are so cool

01:25:52 and like like your play you're you're creating something that actually like even with the slight

01:25:59 prompts they they can operate like humans already in terms of thought.

01:26:06 It's so powerful.

01:26:08 Yes.

01:26:13 And then you have still the most stupid bugs all the time where I'm currently beta testing the app of a team of mine,

01:26:19 a portfolio company of mine,

01:26:20 and they're doing a generative AI game engine,

01:26:24 which is actually incredibly impressive what it can do like it creates 3D assets and everything but it has the

01:26:30 funniest bugs so for example today I said that my player character should be an archer

01:26:36 and the model created was an arch oh my god like a like an arch like an architectural like building arch you know and then all of a sudden you see like an arch? Yeah, an arch. Like an architectural building arch.

01:26:46 And then all of a sudden you see an arch running around shooting laser balls.

01:26:49 You know what I mean?

01:26:49 I feel like that's better than an actual archer.

01:26:52 I'm not disagreeing.

01:26:53 I'm not disagreeing.

01:26:54 But that's the kind of bugs you run into.

01:26:58 For the last five minutes, what would be useful?

01:27:00 Do you want to talk fundraising?

01:27:01 Because one thing you have versus Omer is you're very, very unique, right?

01:27:07 So I assume you had a bunch of people reaching out to you.

01:27:11 What's the general vibe of investors?

01:27:12 Don't feel free to disclose anything that you don't want to disclose.

01:27:16 Just what's the general vibe?

01:27:17 Is it something investors dig or not?

01:27:19 What do you think?

01:27:20 Yeah.

01:27:21 I would say, so the background is I released the I

01:27:26 announced this publicly like three days ago. Oh, well, not

01:27:29 at all. expecting the kind of reaction I was just like quietly

01:27:34 releasing whatever I just thought that this was a cool

01:27:37 idea that needed to be out in the world and have over 1200

01:27:41 people on the waitlist now ready to use it. Yeah, I think the slog of investors, once they see something like that,

01:27:48 they start hitting your DMS and do that. So yeah,

01:27:53 it's, I think, I mean, I, yeah, I don't want to say too much,

01:27:57 but it's been,

01:27:59 it's been fun talking to them and seeing like what,

01:28:03 what they've been saying about it.

01:28:04 This is the, this is the funny thing with fundraising like it's sometimes when you work on products in your

01:28:09 case you didn't experience this maybe just this one but like maybe in the past it's like nothing

01:28:13 nothing nothing nothing nothing some switch everything all at once right dude i've been

01:28:19 building for so long and i've always thought like oh this is the idea oh this is the thing and it's like no

01:28:26 like you're a great founder but like no and then it's like nah here I think my advice there is like

01:28:32 keep pushing to like think bigger and think bigger and then I mean I just had a daughter so for me I

01:28:40 was like that just like shifted my mindset on what's the future I want her to like.

01:28:46 And for me, AI scared the shit out of me sometimes.

01:28:51 And they also excite me.

01:28:53 And I'm like, well, how do I make sure, how do I be part of the team that makes sure that as the future comes, like, there's a scenario, there's a very likely scenario that we like work together harmoniously.

01:29:03 Like with advantages to the human

01:29:06 side which in this case we're getting paid yeah so um also also the beautiful part about like

01:29:13 honestly this is like you tell me when it's getting too dark you know that the vc upside

01:29:19 of like monetizing human enslavement by machines is like huge right like the case here is like that

01:29:28 you have like every human working for machines like running your first system that's like a

01:29:32 huge market right like capitalism is excited here yeah capitalism is i think think about uh think

01:29:40 about every transaction today yeah i think except one i just saw one recently where like an agent bought someone's book on twitter and it was like his first non-human uh purchase

01:29:49 bought some other person's book like oh really okay yes it was an agent that decided oh i want

01:29:57 to buy this book and i think devin the engineering agent did and i was like um i i thought about it and i was like well today a hundred percent of transactions

01:30:09 are human to human five years from now especially with payment it will not be a hundred percent of

01:30:15 transactions are human it'll be into human transactions human to agent transactions like

01:30:20 all of these things will exist and that's the most exciting part about payment is building the

01:30:25 infrastructure this is also like kind of brings us to like the related part here right like we

01:30:30 were talking uh payments in general payments for services by humans and like routing to like the

01:30:36 right marketplaces you know like routing to the mechanic marketplace routing to the influencer

01:30:41 marketplace or whatever um but there's also like an interesting use case where you could

01:30:46 say what if you also build very very smart integrations with amazon and other stuff

01:30:53 so that it's a very very easy for me to share my amazon account with payment but payment makes

01:31:01 sure that this thing doesn't accidentally order stuff i can't explain to my wife you know i mean um and that could be also really really interesting because it's still

01:31:11 in the same area of like i want to give bots the ability to pay for things that they cannot do

01:31:17 themselves you know um but i want to control the spending because i fundamentally don't trust an

01:31:24 llm that might hallucinate that i need like stuff I don't want to explain to my wife.

01:31:29 You know what I mean?

01:31:31 The approval part of purchases for agents, I think that's one thing I hit on a lot, which is the wrapper around the actual figuring out how an agent approaches like doing purchases.

01:31:42 Yep. Makes sense.

01:31:43 I do have to run, Andreas,

01:31:45 too,

01:31:45 but this,

01:31:46 it's always incredible.

01:31:47 Thank you so much for your time,

01:31:49 dude.

01:31:49 This was awesome.

01:31:50 What can the chat do for you?

01:31:51 Like what would be useful for you?

01:31:55 Just if you enjoy it,

01:31:57 go to paymentai.com and sign up for the beta.

01:32:00 And if you want to get connected or like talk,

01:32:03 like follow me on like,

01:32:04 or don't even follow me on like, or don't

01:32:05 even follow me on Twitter. I don't care. Just send me a DM on Twitter. Like, um, it's zero

01:32:10 X Thailand. Cool. Awesome. Uh, thanks so much for your time, dude. Good luck with your fundraise.

01:32:15 If you have any, see you later. Okay. Appreciate it. Boom. Okay. that's crazy.

01:32:29 So I, for one, absolutely welcome our future overlords.

01:32:35 And I'm very, very happy that there will be a service that I can still get some payment when robots take over.

01:32:38 I think this is huge.

01:32:41 And this is kind of like what we had this discussion before, right?

01:32:44 Let me quickly do desktop screen share. So we discussed this before, right? Let me quickly do desktop screen share. So,

01:32:46 we discussed this before, right? Innovation execution credentials. This is kind of

01:32:54 fundamentally new. If you say there's things that we lack even the language for, it's kind of here,

01:32:58 and here is like this is exactly the stuff that's a copy of stuff that exists. It's kind of like, I don't know, here-ish, you know?

01:33:07 Plus it has a really, really good narrative because it connects to things that are creating

01:33:16 like a new what if, you know what I mean?

01:33:18 Like it's a new like why now, new technology and a few other things

01:33:26 and basically

01:33:29 the other good thing here is like in his case

01:33:32 you can easily summarize the narrative

01:33:35 it's like payments for agents, done

01:33:38 like you don't even need the human part

01:33:40 but you can include and say like there are things that are

01:33:44 hard to actually purchase

01:33:46 and like one of them is like human services you know or services done by humans because there are

01:33:52 multiple marketplaces um and so on and so on right um there are multiple marketplaces they are hard

01:34:02 to find they're hard to pay you need the collaboration like you need to like work with these marketplaces so that you can actually pay there you know but

01:34:09 it's in general like also products you know um like for example like the amazon use case

01:34:15 um i think this is extremely interesting and it's a very very clear narrative and the narrative

01:34:21 attaches to a bigger trend to a bigger what now why now to a bigger technology possibility that's opening now and so on and so on and it also connects to like

01:34:29 fundamental like worries we have currently around ai and that's like uh hallucination so like

01:34:36 spending protection and rules you know so not only do you does it connect to like a current

01:34:41 narrative to a current technology it also connects to the thing that we worry about in this space right so it's kind of like hey there's an obvious downside of

01:34:48 lms it's it's hard for them to uh find like to pay for stuff online it's hard for for us to trust him

01:34:56 that they can pay properly and so on and so on so it works very smartly around these limitations so

01:35:03 that's a good narrative and momentum, obviously, apparently

01:35:05 he has. So like, that's good for him. The other risk that here happens, obviously, in my opinion,

01:35:11 at least, is money is important to a lot of people and a lot of businesses, right? And this is the

01:35:18 kind of wisdom, by the way, you're joining here for. The problem you have in his case is it's very likely that open ai

01:35:26 will not only have like a spending for like the the apis you know they might also have

01:35:34 something where their assistant api that's like getting more and more important to them

01:35:39 that this has like wallets you know because it's such a fundamental need and

01:35:48 open ai essentially like if something is like a core need of every customer they have they will

01:35:55 just like try to build this in-house so that's also kind of his biggest risk here the good part

01:36:00 here is realistically it's not there isn't only like open ai there's like dozen other services

01:36:08 and some of them are like um open source you know um if these open source they might not have

01:36:16 like not all of them will have the bandwidth to to build this kind of api and like you know like

01:36:23 this kind of wallets on their own so there's still some value

01:36:26 in somebody just saying you know what i create like a wrapper for all of that and make it possible

01:36:31 and i only focus on making that experience good so that that still makes sense even even though

01:36:39 it's like really really huge so like the biggest opportunity like so what i'm trying to say here

01:36:46 is if i would shape that narrative i would shape it around something bigger not only like human

01:36:52 labor would shape it around like any kind of payments you know of agents focus also like

01:36:57 because like spending protections and so on and so on also like think white here. It can be also maybe virtual credit cards for agents and so on and so on.

01:37:10 Maybe multiple ways to pay.

01:37:18 And that's a very, very big narrative.

01:37:20 That's a very, very big scope.

01:37:23 There's a lot of space to grow into

01:37:26 and he's very early in that so um that makes sense as like an angle towards investors

01:37:33 and the second one i would do here is like um wrapping complexity so you have marketplaces

01:37:41 there are hundreds so even if there will be some people who build this for amazon

01:37:44 marketplaces, there are hundreds. So even if there will be some people who built this for Amazon,

01:37:51 nobody will have the guts to build this for like the Fiverr's, the Upworks, the influencer marketplaces, the mechanic marketplaces and so on and so on in the world, right? And that's kind of

01:37:55 like played for human services, you know? And that's already like a lot of complexity that

01:38:03 needs to be wrapped and OpenAI won't do that.

01:38:06 And even if NLMs get smarter and smarter and smarter, there will be always some edge cases

01:38:11 where you want to cover this.

01:38:13 The thinking here is there is open banking in Europe, for example, but you still need

01:38:21 services like comparable services like play plate to cover all

01:38:26 the edge cases the weird bugs the all the little details you know so even if open ai and the on and

01:38:33 so on and so on will build this for like the big ones amazon and so on and so on you will still

01:38:37 have enough edge cases around you're still enough complexity to wrap here right so this still makes sense the other complexity here is like

01:38:45 different payment methods so i maybe want to uh you know in the background i want to like connect

01:38:52 to my bank i want to use credit cards crypto you know and also like uh is like a backend you know

01:39:05 and to the agent same there right there's a lot of complexity there you have like

01:39:15 pay with bot stripe or whatever they will call it

01:39:20 or like what do you think it will call it actually my thinking is like dev agent or dev

01:39:28 pay or something like this maybe they go back to that their dev payments brand you know um

01:39:35 stripe def i don't know let's see um so you have like this payment option they will have so like um

01:39:44 so you can now pay with credit card but like bots

01:39:46 won't see that they might see like a direct link you know i mean and that can be covered with this

01:39:50 maybe like virtual credit cards maybe maybe there is some sort of like two-way handshake

01:39:58 open source standard he can like play around with sorry not open source oaf like standard that he can play around with

01:40:05 bots have a way to pass on the the payment link directly to this service and the service takes

01:40:12 over the payment there might be like multiple ways to to handle this but what i'm trying to

01:40:16 get around is like get to here is like all of this is like wrapping complexity so So even if OpenAI tries to solve some of that, there's still

01:40:27 enough complexity that makes sense for somebody to properly wrap. And the second part of the

01:40:32 narrative is like the payments for agents, which by itself has so much space to grow

01:40:36 into that you can easily like already now like think where could this go and have a lot of green space to explore and so on and so on.

01:40:48 So yeah, this one is why it's big. Payments for agents, a lot of green space to explore. But this

01:40:56 is true for a lot of AI wrapper products. The problem you usually have is the big players will

01:41:02 just replace you and they just make it as part of the lm but in his case he can like wrap a lot of complexity and none of that is complicated

01:41:11 it isn't complicated to build like an a wrapper for like one marketplace but it's complicated

01:41:16 sorry but it's complex like you have hundreds of those marketplaces and like each one has like one

01:41:22 little weird bug edge case or whatsoever you know and that's

01:41:26 something you're willing to pay a little bit for you know it's not hard to have like a wallet that

01:41:31 can do one or two things you know but if you really really trust that thing you know do you want

01:41:39 you know like do you want somebody to figure out like all the edge cases around bots misunderstanding

01:41:44 what your spending limits are you know i mean you kind you want somebody to figure out, like, all the edge cases around bots misunderstanding what your spending limits are?

01:41:46 You know what I mean?

01:41:46 You kind of want somebody else to do that.

01:41:49 You don't want to, like, own that part of the code.

01:41:52 And same with agents, how agents can pay.

01:41:54 There's, like, multiple ways.

01:41:56 And, like, if the agent faces all of a sudden, like, payment method 5,000, do you want to really build that as well?

01:42:02 You know, like, none of that is complicated.

01:42:04 But, like, together, it's very, very complex.

01:42:07 And that's like an interesting defense against this here.

01:42:11 Cool.

01:42:12 Yeah.

01:42:12 So that's a cool one.

01:42:17 So I think that's it for today.

01:42:20 Let me check.

01:42:22 Pear just joined.

01:42:23 Pear, are you doing your open office hours today or again not?

01:42:27 If you do, let me check. I would recommend everybody comes to you afterwards. No,

01:42:37 he does not. Of course he does not. Pear, every time I'm trying to send you people. Cool. Awesome, guys.

01:42:46 Thanks so much for joining.

01:42:49 Let's wrap it up here.

01:42:54 Let me get my thing back.

01:43:00 Thanks for joining, folks.

01:43:01 That's a wrap.

01:43:08 If you have anything... Wait, Per here are you doing your office hour okay cool cool cool you're doing it in 20 minutes i think right that's your typical time

01:43:14 i'm not i'm not doing 20 minutes of program now no no no no no no no yeah chat is delayed

01:43:20 unfortunately it's like delayed a few few seconds it's kind of boring um

01:43:27 but in 20 minutes uh for anybody who wants to join here the man himself

01:43:35 is live i imagine is streaming um his uh um

01:43:43 open office hours as well with the focus on

01:43:45 open source. It's under

01:43:47 this URL. I highly

01:43:49 recommend everybody going there.

01:43:52 One of the best people when it

01:43:53 comes to commercial open source.

01:43:56 And yeah,

01:43:58 I think it's starting in

01:43:59 20 minutes. If not, Per, you

01:44:01 will let them know. And thanks

01:44:03 everybody else for joining this was a fun

01:44:05 session uh we had people who had their pitch that go viral um and had so much inbound with investors

01:44:12 that they realize that investors aren't really their people and they want to like get first into

01:44:16 a better power position for raising a great decision we had omar who is like currently

01:44:21 middle of the of a small pivot after a very successful launch

01:44:25 trying to figure out how he can move from like a product that starts to work for like small

01:44:30 single player to actually like teams and everything and building a huge huge complex product that's

01:44:36 in a space where there's a lot of competitors you know so how do you build the right not only the

01:44:42 right end product you know with all bells and, but also like what is the right roadmap path to get there?

01:44:49 You know, like what is the step by step things you can build that already early gets like feedback.

01:45:07 uh payments for agents and the ability to actually still get money in the AGI which I think is a good thing because um I I kind of want to get paid by my overlords if they enslave me you know so that's

01:45:13 a good one and in his case uh so different so new and so strongly connected to um um a trend that

01:45:22 it allows like a really really good new narrative you know so all three had like very

01:45:27 very different kind of stages in fundraising first one just like in general getting viral with

01:45:33 investors second one trying to figure out like what is a good positioning but the general with

01:45:38 product but also with investors you know and the third one being like so far outside that it's like really, really

01:45:45 interesting as a bet for investors because it's so different.

01:45:50 You have no idea where it's going.

01:45:51 Right. Cool.

01:45:53 Awesome. That was a fun show.

01:45:56 We will do this again next week.

01:45:57 Thanks so much for joining.

01:46:02 And that's all folks.