Welcome back to the AI Chat podcast. I'm your host, Jaden Schaeffer. Today on the podcast, we're talking about a really interesting company called Grep-Tile. Love the name.
They've just raised $4 million to create an AI-fueled codebase expert. So today on the podcast, we'll be breaking down what exactly this means, what it's doing, and how I think they're a little bit different than some other AI code players in the space. I think this is a really interesting field that's evolving pretty fast. But in my opinion, and if you ask my CTO, my startup, AIbox, not fast enough.
So this is actually founded by a recent Georgia Tech graduate who has been, has essentially been focusing on using AI to help developers understand their code bases, rather than just generating code. So this is kind of a place I think that is perhaps missing or not focused on very much in this general AI space beyond just right this code for me. How do you really understand your code base and how do you help your developers understand the code bases a lot that goes into that? So right now, the startup has an API that developers can connect to their code bases, and they can use it to build custom apps based on a bunch of different AI-driven queries that you do.
So Grep-Tile CEO and co-founder is Daxkupda, and he has described the AI tool as, quote, like, having a highly experienced co-worker who has a deep understanding of your code. I think this is a big deal. Right now with AIbox, we've been onboarding a bunch of new developers over the last couple months. After we did our crowdfunding campaign and we raised a bunch of money, we had the funds to build out our development team.
So we've been hiring a bunch of developers. And one thing that we have always consistently noticed takes a ton of time is when you bring on the new developer, getting them up to date with the code base and everything going on and explaining that, that you essentially have to get another developer to do that. And that costs money because you're paying both developers at the same time for learning. There's kind of this boot up phase.
It's well known in the industry. But I think that there's a significant incentive there to essentially get this boot up down to as little as possible and to not have to use two people to boot someone up. So I think a tool like this that understands your code base and can kind of train and onboard the new people, get them up to speed on how everything's run and written and done, is a huge benefit and literally saves you a ton of money. It's kind of like when you have these AI tools that you can directly say, hey, I used to pay this and now with this AI tool, it's saving me X amount of money.
It gets very tangible. I give you three days of one of my top devs' time is worth this much money. And so if I can save them three days where they don't have to be training someone, I think it makes very tangible monetary impact. So kind of cool.
Right now, the API requires two parameters. So you're connecting the repositories for indexing, right? So you're letting this AM model essentially access all of your code. And then you're adding a natural language query on where you can essentially ask the questions and it can kind of help you understand.
So GraphTel was launched in July, right after the founders thought of the idea at Hackathon, they later joined Y Combinator. I think they were in their winter 2024 batch or maybe I guess they will be in their winter 2024 batch. And Gupta has acknowledged essentially they have made a bunch of different mistakes. He said, quote, first we made the canonical mistake of focusing on attracting investors instead of actual customer needs.
I think this is kind of interesting. This is kind of the big focus was like, how do we get investors into a project instead of like what do customers actually need? Right now, the company has grown to 500 paying customers and they're trying to give software teams with the tools they need to build custom AI developer tools, which are tailored to specific systems. Overall, I think this is a really interesting tool.
I'm excited to see where they go. I'm impressed they've been able to raise this one, which is they have. And the idea I think is definitely super interesting and very valuable. So I'll keep you up to date on how this rolls out in the future.
If you enjoyed the episode, make sure to smash the like button, hit subscribe or you know, review this podcast wherever you listen to your podcast on Apple or Spotify. Really appreciate the reviews and I hope that you have an amazing rest of your day.