The AI-generated summary of my interview with 18-year-old startup founder Arlan Rakhmetzhanov reads like a synopsis for one of the “bildungsroman” novels I studied at university.
“The conversation revolves around a young entrepreneur’s journey and experiences,” transcription app Otter tells me, bringing to mind a genre of coming-of-age stories encompassing everything from David Copperfield to Dune.
Thus far, Rakhmetzahov’s story has followed a familiar narrative: he’s just moved to the big city (London) in search of his fortune (venture capital) and through his tenacity got his first big shot (pre-seed funding to develop his four-month-old AI startup, Nia, from VC firm LocalGlobe).
But what are the odds of a fairytale ending? Rakhmetzhanov reckons his guess is as good as anyone’s. “I’m just a teenager who tried to build stuff without knowing any problems that might arise in the future.”
Nia
Rakhmetzhanov’s product “Nia” is an AI agent that helps developers search and understand their codebase. Rakhmetzhanov started building it in January, “so pissed” by how AI code editors like Cursor and Windsurf “kept misinterpreting things” when he was using them.
“They just suck at coding because they don’t have the mental mode of the whole code like a senior software developer would have,” he says. (Not everyone thinks they suck: OpenAI is reportedly considering acquiring Windsurf for $3bn, while Cursor’s parent company is reportedly in talks to raise at a $10bn valuation.)
Two months ago Rakhmetzhanov dropped out of school, moved from his family home in Kazakhstan to the UK to join startup accelerator Entrepreneur First’s latest cohort and then found himself getting offers from investors within a week of landing. He raised an $850k pre-seed deal led by LocalGlobe, and dropped out again, this time from Entrepreneur First.
With homework permanently scratched off his to-do list, Rakhmetzhanov’s added a few more pressing items: hire a founding engineer; visit San Francisco to get customer feedback; build the AI agent to oversee all others.
In the long run, Rakhmetzhanov wants to build a more comprehensive coding assistant that can execute tasks, understand entire codebases and streamline workflows — just like a senior human software engineer would. Nia could also help developers check if their code base has any security issues, or vulnerabilities — which are currently rife as a result of vibe coding, he says.
None of this will make the software engineer extinct, in his view. “I think humans will be in the loop forever,” says Rakhmetzhanov. “AI programming is going to speed things up — in the latest YC batch, 95% of all code bases are AI generated — but there are still people in the loop. They have to debug everything, because while [AI] might be good at writing code, it’s very hard to maintain, and it might spit out spaghetti code, or wrong functions, and you still have to debug and do all the tests on your own.”
Rakhmetzhanov thinks this approach will give Nia longevity. “I’m capturing not just one tiny problem in some vertical, but a very big problem coming with the rise of vibe-coding AI.”
Build and build
To get there, Rakhmetzhanov wants to talk to a lot more users — and thinks he’ll find them more easily in the Valley.
“Looking at who’s buying my product, they’re all primarily from San Francisco, because they’re developers and startup founders — and they’re all gathered in one place,” he says, adding that he’s considering visiting this summer, and potentially making the move long-term.
“To keep up the momentum, and iterate very fast, I think San Francisco would be a better place [to be based] because there it’s just a builder culture, while London might sometimes distract me. It’s a capital of the world — it has so many restaurants, bars, a lot of people to meet… In California, you just build and build, and you don’t leave your apartment.”
Top of the agenda: “Finding my first 100 power users, getting as much feedback as I can and developing the product features-wise to do even deeper integrations into developer workflows.”
Rakhmetzhanov also plans to hire a founding engineer.
Is he daunted, I wonder, by the prospect of recruitment? “This is my first experience hiring, so I’ve been asking a lot of advice from people from LocalGlobe.”
He’s currently writing a job description, and plans to forego the usual “boring coding interviews”. “I just need someone crazy, young and ambitious.”
He’s also hoping that person can help him sell the product, by “talk[ing] in the same language” to prospective customers.
Tenacious
“Arlan is one of the most driven founders we’ve met,” says Emma Phillips, the partner at LocalGlobe who led the deal.
Rakhmetzhanov, who’s drinking a Diet Coke at 10am in the morning when we speak, has a long CV for someone who was at school full-time until recently.
“I was really obsessed with getting into Stanford, Harvard, MIT, so I built this internal tool for myself, which is an AI-powered database of scholarships and grants, to help me apply to those programmes. I did a couple of them, and my friends started asking for that tool, and I thought, ‘Ok, why not monetise it, and scale it?’ So I started posting about it on Reddit and TikTok and eventually scaled it to 20k users and made my first $1k.”
Rakhmetzhanov also cold-emailed over 300 professors in his bid to make it to the US. Only one replied — Ilya Strebulaev, professor of finance at Stanford, who offered him a remote internship as a research assistant analysing VC trends last October.
He’s applied to YC twice (“I got rejected. I’m applying again, because I’m not stopping”), Sequoia’s Arc accelerator (“didn’t hear back”) and A16z’s Speedrun.
In February, after two weeks of building Nia, Rakhmetzhanov launched it on Product Hunt where it made top three product of the day. Buoyed by the positive feedback, he cold-emailed a bunch of YC founders, including James He, founder of AI startup Artificial Societies, asking them to check out Nia and share advice on how to launch the product.
“[He] really liked it, and immediately offered me an angel investment while I was back in Kazakhstan. I was really shocked.”
It was He who then introduced Rakhmetzhanov to LocalGlobe, and some of his other investors, including Gideon Valkin from Andrea Ventures and Ramzi Rafih from No Label Ventures.
“Before coming here I heard from people that VCs suck. They just want to give you money and quit. But that’s not the case, at least in London.”
Nobody seemed worried about his youth, he says. “When I told them I dropped out, they probably saw it as a sign of courage; it just shows that you’re determined to build your company and go further.”
Make something that people want
Rakhmetzhanov has no plans to raise any more funding soon — and says if a big-name VC were to pre-empt a round, he’d think very carefully about whether to take the money. “From an early age, my mum taught me not to treat money as a thing to buy cars, or expensive watches, but as something to invest in myself and in my vision. I see money as a resource.”
For now, he just wants to build, and follow the same advice he’s been following for some time, Y Combinator’s slogan: “Make something that people want.”
Rakhmetzhanov says: “It’s stuck with me for years now.”
#Meet #18yearold #dropout #building #agent #rule