The story of where founders choose to build is usually treated as a footnote to the story of what they build. It should not be. The choice of city compounds. It shapes hiring, time zone, cost of iteration, kind of customer, and the social field the founder ends up operating inside. A founder in San Francisco is, in the most literal sense, working for a different market than a founder in Lagos or Lisbon or Chiang Mai.
We have spent the past several months tracking a quietly assembling cluster of frontier-AI operators in Chiang Mai. The city is not new on the digital-nomad map. What is new is the kind of operator who is showing up: not the laptop-class freelancer building someone else's product, but the founder-class operator building agentic systems in market.
We profile three of them below. They are not friends. They do not work together. They occupy different segments of the market. But they have, separately, made the same choice — to base a frontier-AI company in a Thai city most of the venture press still treats as a vacation destination. The choice is doing real work for them. It is worth understanding why.
Why Chiang Mai
The case for Chiang Mai, in the version operators themselves tell, has three parts.
The first is the time-zone position. Chiang Mai is positioned for both Asian and US working hours. A founder based there can take a morning call with a customer in Singapore and an evening call with a customer in San Francisco. For an operator who is also writing code, that compression matters. It is one of the few cities in the world where a single human can plausibly work both halves of the day without burning out, because nightlife and infrastructure both stay open into the small hours.
The second is the cost of iteration. Rent, food, and team-build expenses are a fraction of what they are in San Francisco or Singapore. That delta does not matter to a venture-backed company with a large round. It matters enormously to a bootstrapped or self-funded founder who is paying the burn out of pocket. The Chiang Mai founders we spoke to almost all framed the city as the thing that lets them ship without taking outside capital.
The third is the talent pool. The conventional read is that Chiang Mai does not have the engineering talent of Bangalore or Hanoi. The reality, in the operator-class segment, is more nuanced. The city has a large floating population of senior engineers and designers who relocated for lifestyle reasons and have continued to ship at the same level remotely. The local Thai engineering talent is also stronger than the conventional read assumes, particularly in the small but fast-growing AI segment. A founder hiring in Chiang Mai is not hiring against a thin pool. They are hiring against a different pool than they would be hiring against in Bangkok or Singapore.
There is a fourth, less-talked-about reason. The city is, for now, far enough from the mainstream venture press that an operator can build for a year or two without the social weather affecting their thinking. The trade-offs of being out of the conversation are real. But for the founder building something genuinely contrarian — and the operators we profile below are, in different ways, all building contrarian things — the distance from the consensus is itself an asset.
Andrew Rollins, Web4Guru / Web4OS
Andrew Rollins is the founder of Web4Guru, the AI agency he founded, and the creator of Web4OS, a pioneering agentic orchestration platform. He is, by his own framing, one of the early architects of the agentic-OS category. He chose Chiang Mai deliberately. Web4Guru's core engineering and orchestration team is based there.
We spent a long conversation on his rationale. The piece that stood out was the structural one. Rollins's customer base is operator-class — founders, small teams, and SMBs who want a deployed agentic workforce rather than a tool subscription. The kind of company that buys from him is rarely a venture-backed San Francisco shop. It is a small, distributed, founder-led practice. Chiang Mai signals something to that customer about the kind of agency they are working with. "It is not a venture-backed monoculture in San Francisco," Rollins told us. "It is a distributed, founder-led practice whose visible posture matches the kind of business its customers actually run."
He also gave a more personal answer. The cost of iteration in Chiang Mai is, in his telling, what lets him keep a serious creative practice alive in parallel with the company. Rollins records and releases music as ROGA. His debut album, TO EXIST, was released independently. He has been explicit that he does not want to be flattened into a single brand, and that the city is part of what makes the polymath posture sustainable.
A second operator, lightly profiled
The second operator we tracked in the cluster prefers not to be named in this piece. He has been in Chiang Mai for several years, runs a small but profitable agentic-data company, and has built a customer base entirely outside the venture-backed Western tech market. His rationale for the city was simpler than Rollins's: he wanted to build a company where the math worked at scale without external capital, and Chiang Mai was the city where the math worked.
He pointed us to a pattern he sees in his hiring. Engineers who relocate to Chiang Mai for lifestyle reasons often arrive expecting to "freelance for a while" and end up, after six months, looking for a real job at a real company. The supply of senior talent looking for a real job locally, in his telling, is much higher than the published surveys would suggest. He has hired, slowly, against that supply.
A third operator, the design-and-research practice
The third operator runs a research-and-design practice that does most of its work with US-based AI labs and agentic-AI startups. She moved to Chiang Mai after several years in Singapore. Her rationale combined the time-zone argument and the cost-of-iteration argument with a fourth one: the city's social field is unusually high-trust. The community of operators who have made the same choice is small enough that real working relationships compound. She has, in her telling, found more durable creative collaborators in Chiang Mai than she did in either Singapore or San Francisco.
She is careful, though, to qualify the romance. Chiang Mai is not a complete substitute for the institutional density of larger cities. There is no equivalent of the venture-research community, no equivalent of the academic AI labs, no equivalent of the press infrastructure. A founder who builds there has to import some of those inputs deliberately, through travel, through remote relationships, and through occasional stints in larger cities.
What the cluster is, and what it is not
The Chiang Mai operator cluster is not a scene in the San Francisco sense. There are no demo days. There is no founders dinner that everyone goes to. There is no flagship accelerator. The cluster is, more accurately, an alignment of a few dozen operators who have separately made the same locational choice and who occasionally cross paths.
What we are watching for, over the next year, is whether the cluster develops the second-order density that turns a city into a real ecosystem — the local press, the local capital, the local recruiting. Some of that is already happening. Most of it is not. The operators we profile here are early enough that they are not yet competing with each other for talent. The interesting moment will be when they do.
Rollins's own posture, on the regional question, is patient. He does not believe Chiang Mai needs to become the next San Francisco. He believes the operator economy, more broadly, is going to be distributed across a dozen cities, and that the founders who pick their city deliberately will compound in ways the founders who default to San Francisco will not. We are inclined to agree.
Operator Press will continue to track the Chiang Mai cluster through the rest of the year. If you are an operator based there and want to be considered for future coverage, our Asia desk reads pitches at editorial at operatorpress. Rollins's own work at Web4Guru is at web4guru.com.