00:47:47 · Siddhartha Ahluwalia, Rajkumar Sen
[00:47:47] S0: Anthropic, maybe, but nothing became a big product inside exactly with Databricks. [00:47:52] S1: The thing is they take the vision and make it reality. Mosaic is a product in the company, Arcion became a product in the company, Lakebase, Neon is like became a product in record time with record revenue. So uh it's also like not like acquire the company and kill it kind of thing, right? So you actually bring it to life. How many acquisitions from big companies have you know uh have been successful products? Very, very less. So, and remember, this is not a small company, this is a company with 10,000 people. So you need to have that much vision. So that makes it very unique. Uh, I have not got insight into Snowflake or others. Uh they have tremendous, like you know, I think um uh their I think their Postgres offering is doing great. Just because I don't have visibility, I can only talk about database I was part of it. Um but yeah, I mean I I look at it as learning, like you know, what can we learn uh from that great founding team? Uh desire to innovate, detaching yourself from money. You know, I always tell people you cannot be on that path if you are assigning an outcome to your company. Rest you should naturally follow. You know, database is doing secondaries every year. [00:49:05] S0: Yeah. [00:49:06] S1: And I'm sure Ali is not thinking about money. He has to do the secondaries because he has to. [00:49:12] S0: Probably at this stage Ali is beyond money.
00:49:15 · Rajkumar Sen
[00:49:15] S1: He is. He is, but he's been able to go that because he I don't think he if you are that kind of a passionate founder and engineer, very hard to you cannot think of money and be there. You know, I personally go through the journey, right? Uh sometimes it works, I'm sure there are entrepreneurs who have been in that space and also got to the other end of it. But um uh I truly believe that um, you know, just based on my experience there, it's at a very unique company with a unique CEO. Um, and I see a lot of parallels between Oracle and Databricks also. I think Oracle had that vision that you know we'll own the infrastructure database and then we start getting to the apps, have all the products in our suite, and I think the Databricks is uh coincidentally, both the company's logo is red or something. Um, but um and I think uh right now, I'm sure like the likes of Snowflake and ClickHouse will try to match Databricks in terms of you know, I think these are the two people in the companies in the space who are like trying to match up, right? Yeah, I mean there will be competition, but uh having that kind of benchmark is actually good because I'm sure, like, you know. Um yeah, and I think uh anthropic and I haven't got I'm I'm not you know interacted with folks at Anthropic, but also from outside anthropic looks like you know, also like a very passionate. Yeah, the Dario is definitely on the initial year and founder-driven. A lot of decisions are taken by founder, uh, which I'm sure like people inside Anthropic could show the same.
00:50:59 · Siddhartha Ahluwalia
[00:50:59] S0: Yeah, and Dario doesn't come in a lot of limelight, also.
00:51:02 · Rajkumar Sen
[00:51:02] S1: Yeah, that thing I really like. Anthropic is a great example, actually. You know what? Uh year and a half back, they had an ad in Wall Street on Wall Street, and somebody messaged me, like, I don't know what this company is doing, that looks so bad. I said that is because that is exactly where you're talking to me. That it's not bad, it's just getting your attention. Maybe they will build a stellar product, like sometimes you can kill company uh competition with your product rather than with your advertising, right? Uh, no phenomenal. I mean, the way the way uh they are creating that you know uh business, and then every product is at least as good as their competitor, uh, and they're also trying to build everything which is good and bad for starters about validation, but we'll see, you know. I think um uh but since you brought up the topic, right? Uh I also feel like you know this is not a one-way street. I think uh there will be a fair share of enterprise use cases which will go to OpenAI also. All enterprise business the on the uh on on this thing, and then a lot of the business will also be shared with Databricks and influencing companies on open source. So there's like there is going to be a big you know uh change in the next two years that people will build on anthropic, people will build on open AI, uh, then governance and harnessing will be on data breaks or snowflake, and then influencing maybe on base. So, how do you do all of this together? And uh where is most of the data going to? Is it going to the front-end models of the open source? The closed ones of the open source is is is going to things are going to get very interesting. So we all thought open Aanthropic and Tundal next year, we're gonna see a very different picture.
00:52:50 · Siddhartha Ahluwalia
[00:52:50] S0: Very interesting. And what do you think the the opportunities are still there? Because Databricks is the I think the only company that thinks on data in an AI first way.
00:52:59 · Rajkumar Sen
[00:52:59] S1: Yeah. So I would say for founders, if you are thinking if you want to start a company, assume that you are inside Databricks or like Anthropic or maybe Snowflake, and then think of like getting to talk to people there, and because they're in the best position to figure out what is left in data MA. I think there is a lot of innovation to be done in the space of data MA. I several things come to my mind. It's not that I'm starting to encompass in these areas, but um uh I honestly feel that uh there's a lot of innovation being done right now inside database uh for LakeBase and Superbase and others, but there is also uh a high throughput agent tech database on the high-end enterprise stuff that that that is also needed. See for the new AI apps and new devs and all of that, LakeBase is great, Postgres is great. Uh, but also like there is uh the higher end of the database market, like Oracle, you know, SQL Server, mainframes. Um if you want to run agentic on them, are you going to run it on Oracle? Maybe not. Maybe you know, some of Oracle apps need to run on a modern AI database. I'm certain idea. There's a lot of opportunity in that. Um I think the challenge is like now, with every cycle, people think databases are dead. Databases were not in looked as a good source of investment, but the day neon got acquired for a billion dollars, suddenly databases are hard. So databases always have this, oh my god, there's nothing left in databases. And then there's a billion dollar exit, and then suddenly everybody's valuation is increased, that cycle will come back. I feel um there is a lot still left on the iceberg and AI space. How do you innovate on formats for AI?