00:14:24 · Siddhartha Ahluwalia, Rajkumar Sen, S2
[00:14:24] S0: This is 10K. You started the company in 2017. [00:14:27] S1: Yeah, 18, yeah. So I said, you know, if you bring value, yeah, I obviously the your buyer is going to want to pay you zero dollars, but you need to have the conviction and at some point say no. [00:14:40] S2: Yeah. [00:14:41] S0: Um even in Bay Area, the buyer wants to pay you zero dollars everywhere. [00:14:46] S1: Uh, because that's the nature of the buyer. I thought that is only in India. Oh no, no, no. Especially if you are a small company, yeah, it it's a very prevalent thing because you want to do escrow, you want to do this. It was a big battle for us to push back and you know, uh say this is the last thing. And how much did they initially quote? [00:15:06] S0: $10,000.
00:15:06 · Siddhartha Ahluwalia, Rajkumar Sen
[00:15:06] S0: $10,000. [00:15:08] S1: I think they said we can pay you like $10k or something. Okay, and then we started with $100. I didn't know like because you know, I I I realize that is very essential for business. Yeah, so they can't survive without you. So, but obviously we cannot charge a million or 100k because you know it's also how we have to prove it out. Yeah, you know. So we said we're gonna, this is more of a design partner, they're gonna help us build the product also. But then I don't want to stretch too thin because if I take on this customer, I'll have a couple of engineers busy and all of it. It took, it took, it took a long time with procurement. Uh no, closing took a month only, like for from right from the first day. From the first uh initiate uh conversation to closing took five and a half months. Yeah, uh, but then three months was to build the connector, and you know, we first built a um platform to move data in first infrastructure, then we started building connectors on top of it so that you know whether you build MongoDB or Cassandra doesn't matter. Uh so I always tell founders that you know if you believe that the software has value or your services a value, even if it's the first customer, do not sell for cheap. Because that sets a precedent for the next one. Yes. If you start selling for $10,000, next one can you sell for $15 because in your mind you already accepted defeat. So we had um the second customer actually at 100k. [00:16:31] S0: And who was the second customer? [00:16:33] S1: Um it was Dell. [00:16:35] S0: Okay.
00:16:35 · Siddhartha Ahluwalia, Rajkumar Sen
[00:16:35] S0: Okay. [00:16:35] S1: Yeah. So there's uh similar use case, not on MongoDB, but from a uh they wanted to do real-time replication from one of the largest warehouses uh to a modern like platform for real-time analytics. So, you know, as I said, like there is no we didn't come at a pattern because uh the pattern I actually wanted was from on-prem through all app systems. And I was trying to go to this big data warehousing company said, hey, why are you pushing files and recomputing the whole thing every day? Because the source table probably changed only one row, but you're but everybody is making money, you know. It's like if you're spinning up a big compute to redo the batch, everyone's you will make money. But uh, I think that twist came when Databricks launched Delta Lake, and then I think people are working on similar things and iceberg because Netflix was like, you know, event design is the future. Uh, and Databricks came up with a paper Delta Lake. They said no, no, no, no, no, you know, you need to do CDC into Data Warehouse, and they'd also give the opportunity to, you know, make the format open source. I think Delta Lake and Iceberg were tremendous because they launched at the same time, right? [00:17:47] S0: 2018, 2018.
00:17:47 · Siddhartha Ahluwalia, Rajkumar Sen
[00:17:47] S0: 2018, 2018. [00:17:48] S1: And uh, I always say, like, to you know, all Databricks folks and ex-Arcion folks, that if that's what the market needs to also appreciate what we're doing and come together. If there are no Delta Lake and Iceberg, the exit for Arcion would have been 18 million because we had an offer for 18 million from one of the other database companies uh for migrations. You didn't have the replication use case and without Delta Lake or Iceberg. Yeah. Uh if you had run out of money before Delta Lake or Iceberg had come, we would have been bankrupt. [00:18:18] S0: So tell me about the the fundraising journey in Arcion. How was the first round? What a prox dilution or valuation you did for the first subsequent round. [00:18:28] S1: Yeah, it was uh very difficult, but also I got like uh investors who believed. [00:18:36] S0: Yeah. [00:18:37] S1: Um initially we had uh some ancient money from some ex uh Facebook uh gang uh who knew me from MemSQL days as a director of engineering doing stuff. So look at this, we put some money in him. But uh obviously, you know, I had a set of people who were with me, but then when they saw fundraising challenges, they went back to their own. Yeah, yeah, they all let you know they're I still super good friends with them, but naturally, you know. [00:19:07] S0: Yeah, but living in Bay Area is hard.
00:19:07 · Siddhartha Ahluwalia, Rajkumar Sen
[00:19:07] S0: Yeah, but living in Bay Area is hard. [00:19:10] S1: If you're already uh stable and all, then you know, I mean it's a then they're not not able to raise that much. Um but um I had a good fortune of uh through one of them actually, uh one of the uh founders, co-founders meet uh Alok Nandan, who at that point of time was in emergent ventures Ventam at first race, you know him very well. Um Alok actually had stayed at Azure and Salesforce, and he also struggled with that problem while he was at Salesforce. So he said, I know this problem for machine learning, it's needed for you know triggering new jobs and all. Uh, but maybe the problem exists for databases also. So I think, and then um got the opportunity to talk to Uncle at Emergent. They led the initial um seed round, and uh obviously they were also a small fund at that time. Yeah, that gave me the initial capital to build a team. I uh hired somebody in India um uh and Europe, uh, and one person here, so we scrappy team. Yeah, uh, but then you know, yeah, it's um it was quite a journey, you know. [00:20:24] S0: Um you said you came to shutting down multiple times. Yeah, yeah, yeah.
00:20:24 · Siddhartha Ahluwalia, Rajkumar Sen
[00:20:24] S0: Um you said you came to shutting down multiple times. Yeah, yeah, yeah. [00:20:28] S1: That I think um probably is true for many startups. We had that uh situation uh a couple of times because uh COVID also impacted us badly because spending was so initially, whatever round you raised, let's say less than a million dollars, you would have run out of it in 18 months. Yeah, yeah, very soon. And also um because, as I said, right, we got bits and pieces, right? Some MongoDB, some Teradata to Postgres, some migration, some Cassandra to somewhere. But there is no theme, yeah, which is also getting hard to fundraise. You know, we talked about one of your LPs, right? Which extensively talk to them. People are understanding that this is this is needed in future at some point. This has to be there. Without it, economics of data movement would scale for. But then they were not seeing the value right away because you know there's no customers doing that. Um, and so fundraising was hard. Uh obviously, we talked to Informatica once and others uh when we were about to, you know, uh and I was on H1 also. Okay. Uh so as you mentioned about dilution, right? Like, you know, because if you're on H1 as an employee, you cannot have more than X percent uh of the company. So I obviously started the company with one half of what a regular founder would have, but you know, that's you know, I and who had the other? [00:22:00] S0: Half investors.