Algora | Algora TV

Tigris Data Founder Interview Highlights

Algora

Watching now

00:00 I have been doing databases and data infrastructure for more than 17 years.

00:04 I spent almost six years at Uber and I was head of storage infrastructure there.

00:08 We built our own distributed database called DocStore.

00:11 We had built our own global replication service to replicate the data globally and then eventing

00:16 framework.

00:16 Working at scale has its own challenges and then running a platform with a critical database

00:22 platform for a company like Uber has, you know, has its own

00:25 unique challenges and learning opportunity. Left Uber in 2021. And with two of my colleagues from

00:32 Uber, we started Tigris in December, 2021. Four corrective stakes of a database around, you know,

00:37 security, high availability, performance, scalability that have to be there. We spent a

00:43 year and a half, and I would say that we, that's a record time to build a database. So the first commit we did was in the open source

00:50 repository with an Apache 2 license. And that's how we have been working. Everything that we have

00:54 done is in the open. Everything that we have done is available, free to use as open source.

00:59 A majority of the development initially was driven by us. And now we're able to provide

01:04 better support to open source contributors. And now we're able to provide better support

01:05 to open source contributors.

01:06 And we have started to see those contributions

01:09 come in over the last month or so

01:11 since we did the launch.

01:14 Definitely a great period.

01:15 I'm very excited about the production workload

01:18 of production traffic.

01:20 And it's an amazing milestone to hit.

01:24 We have modern cloud native architecture. We have modern cloud-native architecture.

01:26 We have a desegregated architecture

01:28 where storage and compute is desegregated.

01:31 And that is why we are more efficient,

01:34 more scalable as compared to DynamoDB and MongoDB.

01:37 Actually, our users tell us that it's three times faster

01:40 to get started with Tigris as compared to MongoDB.

01:43 We tend to think that

01:45 open source is free, but if open source were free, then how would people make a living out of it?

01:52 There has to be a sustainable business that needs to be built around open source.

01:56 So the way we are monetizing the product is by offering a hosted version of the product.

02:00 What we did differently was that not only are we building the core product in open,

02:04 of the product. What we did differently was that not only are we building the core product in open,

02:11 but we also worked on the orchestration platform layer simultaneously. The decision to do that was based on experience. So experience tells that this is something that has to be done.

02:16 When we say that it's a serverless platform and that it makes developer experience

02:21 really simple and easy to use, if we then ask the developer to deploy

02:25 it themselves, then we are not really showcasing the simplicity of the platform, right?

02:32 So from that perspective, it always made sense to make a hosted version available.

02:37 And then from enterprise use case perspective, definitely they do need orchestration layer

02:43 so that they can deploy it or they can use it as

02:46 their internal platform. Absolutely. Thanks for highlighting this. Very helpful. And I think we

02:52 will take note how this approach makes more sense and it's more demanding to ship and to ship fast.

02:58 Yeah, definitely more demanding. Twice the work. In terms of go-to-market our strategy is mainly dependent on you know

03:05 building awareness um and that's done through marketing put it in front of the user and that's

03:10 what marketing is essentially but then when you put in front of the user and user it's solving a

03:15 problem for the user the product needs to show that that you know it's solving your problem it

03:19 has the right experience that you expect out of it absolutely and for a developer tool an

03:24 infrastructure specifically,

03:25 that's the only way that makes sense, I think.

03:28 To see other people around you

03:30 experimenting with the technology,

03:32 using it to drive awareness.

03:34 And I actually saw on your Twitter earlier today

03:37 how this developer with a big following

03:39 who does content teased

03:41 that he will be using Tigris in a future video.

03:45 So I was actually curious to ask you, like, how did this come about?

03:49 Did you reach out?

03:50 Did he reach out?

03:51 How did this come about?

03:53 Yeah.

03:55 Yes, I reached out because, you know, we have to build awareness.

03:59 And there are some amazing people like Jason who are helping the community,

04:02 teaching how cool the technology is, the problem it's solving, how it is making database access used in storage simple and easy.

04:09 You know, we have been using the bounty program, open source contributions and making sure they're

04:16 paid so that what you're doing, you're actually, you know, getting paid for it. And that's the way

04:23 to make open source sustainable,

04:27 to not expect people to do things for free.

04:29 Helps us to improve the product,

04:31 helps us to get more people in,

04:33 to contribute in ways or areas that we are not able to, right?

04:35 And get a different perspective.

04:37 So it has been great for us.

04:40 And I wish the same can be done

04:42 about the content as well.

04:43 Why not get the community involved?

04:45 Provides a more structured approach to, you know, building out content and showcasing

04:50 how a certain technology can be used.

04:52 Thank you.

04:53 Thank you so much.

04:53 And then what I like specifically about this idea of, you know, content boundaries, because

04:57 the reward can be immediate, like even before building a big following in a big channel,

05:01 that sort of like, I think opens it up to more people giving it a try and you know exploring their talents so we're definitely gonna run this experiment so thank

05:10 you so much is there some additional advice or maybe a mistake to avoid from from your experience

05:14 so far that we can share with other founders um mistake to avoid um yeah i think that uh especially

05:23 you know people coming from engineering background like myself, we tend to think a lot about the architecture and engineering.

05:31 And I think it's more important to think about the problem that is being solved.

05:35 What problem I'm solving and why am I solving it?

05:39 Who is my user going to be?

05:40 Those are really important questions to ask.

05:44 And not purely be feature driven,

05:48 be value driven. What value am I bringing? And if I'm not able to answer a question about what

05:53 value I'm bringing, then I should not be focusing on the future. So that's a matter of priority,

05:58 but I think that this is something critical that needs to be taught up front we have to you know just remind

06:06 ourselves that we have to adapt and do what is needed by the business if the business look i

06:12 never did marketing before uh yes i wrote blog posts in the past but i never did marketing before

06:17 i never did sales call with users before i never did feedback calls with, you know, customers outside of the company. Right. So this is all new for me, but I really enjoyed to start up a company and, you know, and to

06:30 be able to start a company with my friends, that was, that doubles the joy.

06:36 But yeah, it's an amazing experience and loving experience and, you know, being able to, you

06:41 know, bring value.

06:42 That's, that's all I want.

06:44 Yeah.

Library