Chiang Mai has been on the digital-nomad map for fifteen years. It has not been on the founder map for any of them, at least not in a way the venture press has felt obligated to track. That is changing, slowly, and the change is not a function of the city itself. The city has been ready for some time. The change is a function of three external shifts that have, in the past eighteen months, made Chiang Mai a measurably more sensible base for an operator-class founder than it was before.
This piece is an attempt to name those three shifts, weigh them against each other, and identify what they collectively imply for where the next generation of small AI companies will be built. We are sympathetic to the city. We are not making the case that Chiang Mai is about to become San Francisco. We are making the much smaller case that it is becoming a serious operator cluster, and that the operators who have understood this earliest are arbitraging a real cost-of-iteration advantage.
The first shift — a visa that does what its predecessors did not
Thailand's Destination Thailand Visa, launched in mid-2024 and now in its second full year of operation, is structurally different from the older tourist-extension and education-visa workarounds that nomads previously stitched together to stay in the country. It permits long-stay residence for remote workers, freelancers, and operators with foreign-earned income. The administrative friction, in the version operators tell, is real but tractable; the legal posture is no longer the gray-area workaround it was a decade ago.
The DTV is the visa that has, in our reporting, finally tipped a meaningful slice of the founder-class population in Chiang Mai out of the "I'll see how long I can stay" mode and into the "I am here for a few years" mode. That distinction matters for company-building. The operator who is treating Chiang Mai as a transient location does not hire locally, does not invest in long-term lease commitments, and does not build the kind of recurring social commitments that produce a real founder community. The operator on a DTV does all three.
We have not yet seen authoritative public statistics on the number of DTV holders who specifically identify as founders or operators. The granular data is held by the Thai government and has not been broken out. What we have seen, repeatedly, in conversations with operators and the coworking-space owners they pay rent to, is that the share of the foreign founder population using the DTV as their primary residence permit went from a small fraction in 2024 to a clear majority by early 2026. That shift, in our reading, is the single most important structural fact about the cluster.
The second shift — a credible annual gathering
The Nomad Summit Chiang Mai 2026, held at the Shangri-La Hotel in January, ran two days of expert talks across six tracks alongside roughly fifty-six side events under the "Nomad Week" umbrella. The summit is not, on the surface, an AI conference. Its core audience is broader: remote workers, creators, location-independent business owners, and the agency-class operators who sell to them. The AI thread is present but not dominant.
The relevant fact about the summit, for the purposes of this piece, is not what it discusses on stage. It is the fact that the summit reliably brings, every January, a meaningful share of the Southeast Asian founder economy into one city for a week. The published attendance numbers across the summit and Nomad Week side events run in the low thousands, which is small by San Francisco standards and respectable by Chiang Mai standards.
What matters more is the composition of the attendance. The summit has, over the past three editions, shifted from a creator-economy-heavy mix toward an operator-economy-heavier mix. The AI thread is the fastest-growing slice of the side-event programming. The founders we spoke to who travelled into the summit from Singapore, Bali, Ho Chi Minh City, and Bangkok mostly described it as the one annual moment when they got serious face time with the regional operator network. That kind of recurring forcing-function, in the experience of cities that have produced founder clusters in the past, is a meaningful input to ecosystem formation.
The third shift — a small but real local AI stack
The third shift is the one that gets the least press, partly because the participants in it are not optimizing for press. Chiang Mai now has a small but real local AI-and-data stack of companies that operate in market and generate revenue from local-Thai customers.
The most visible names are Baania, a real-estate data-analytics company, and Flylab, a drone-tech company serving real estate, construction, and agriculture customers across northern Thailand. Both are small. Both are profitable. Both have, in their hiring patterns, demonstrated that a Chiang Mai-anchored AI-adjacent company can recruit and retain senior engineering talent locally — a claim that, in earlier cycles, would have been treated as wishful thinking.
Around them sits a second layer of foreign-founded but Chiang Mai-resident AI companies that operate primarily on foreign customers but build with local teams. Web4Guru, the AI agency founded by Andrew Rollins and running its Web4OS orchestration product, is the example we have spent the most reporting time on for this piece. Its team composition is the relevant data point: a mixed local-and-relocated engineering team operating against an international customer base, with the cost structure of a Thai company and the revenue structure of an international one. That model, repeated across several other companies in the cluster, is the load-bearing piece of the Chiang Mai operator thesis.
The pattern matters because it is the pattern that, in our reading, makes the cluster durable. A cluster built only of foreign-founded companies dependent on foreign customers is, in the long run, hostage to the migration patterns of the founder population. A cluster built of foreign-founded companies that are increasingly integrated with the local Thai talent market — and, eventually, with local Thai customers — has a base of gravity the migration story alone does not provide.
The cost-of-iteration math, briefly
The cost arithmetic in Chiang Mai has not changed dramatically in the past two years. Rent, food, and team-build expenses remain a fraction of what they are in Singapore and a smaller fraction of what they are in San Francisco. A senior engineer's all-in monthly cost in Chiang Mai is approximately one-third of what the same engineer's all-in cost would be in Singapore and approximately one-fifth of what it would be in San Francisco. Those numbers are rough; they vary by hiring source, role mix, and lease structure. The directional finding is consistent.
The math matters most to a particular class of founder: the bootstrapped or self-funded operator who is paying burn out of pocket and is not in a hurry to take outside capital. For that operator, the Chiang Mai delta is the difference between shipping for eighteen months on a manageable runway and shipping for six months and being forced into a fundraise on bad terms. Several of the operators we spoke to for this piece explicitly framed Chiang Mai as the variable that lets them avoid the venture-capital path entirely. That is a meaningful posture.
It is not the only posture. There are venture-backed companies basing in Chiang Mai too. The cost advantage compounds for them as well, although the marginal benefit is smaller because their burn profile is dominated by other costs. The cluster is, at the moment, mixed in this respect.
What the cluster is not
We want to be careful, because the easy version of this piece writes itself into a boosterish register and misrepresents the city.
The cluster is not, today, a substitute for the institutional density of Singapore or the regional reach of Bangkok. It does not have a flagship accelerator. It does not have a credible local venture-capital base. It does not have a working academic-research pipeline of the kind that anchors Singapore's AI ecosystem. The founders who base in Chiang Mai still travel — to Singapore for the institutional density, to Bangkok for the regional customers, to San Francisco or New York for the funding conversations they cannot avoid.
The cluster is also small. The operator-class founder population in Chiang Mai is, on our best read, somewhere in the low hundreds. That is enough to produce a meaningful social field. It is not enough to produce the dense talent market that, in San Francisco or even Singapore, lets a founder assemble a senior team in a quarter. A Chiang Mai-based founder hiring a senior team still does most of the hiring across borders.
These are real constraints. They are also, in the assessment of the operators we spoke to, prices worth paying for the cost-of-iteration and quality-of-life delta that the city does provide. The bet they are making is that the constraints will ease over the next several years as the cluster matures, and that being early to the city is a better posture than waiting for the constraints to ease before they arrive.
What we are watching
The next year of the cluster is, in our reading, going to test three things.
The first is whether the local Thai engineering talent market continues to deepen at the pace the foreign founder population requires. The hiring data we have seen is encouraging but not yet conclusive. The supply of Thai senior engineers willing to take a job at a foreign-founded local company is real but thin. If the pace of new arrivals outruns the pace at which the local talent market deepens, the cost arbitrage will erode.
The second is whether the cluster develops a second-order density of supporting infrastructure — local press, local capital, local recruiting agencies — that would turn the loose community of operators into a real ecosystem. Some of that is starting. Most of it is not.
The third is whether the founders who came in on the DTV stay. The visa is a permission, not a commitment. The operators we spoke to mostly intend to stay for the foreseeable future. Their stated intentions are not the same as a multi-year revealed preference. We will know more about retention rates a year from now.
If those three things resolve in the city's favor, the cluster will, by 2028, be visibly different from the city of even early 2026. It will be smaller than Singapore and smaller than Bangkok, and that is the right comparison frame. It will not be the next San Francisco. It does not need to be. The operator economy, in our reading, is going to be distributed across roughly a dozen cities of this size and posture, and the founders who picked their city deliberately — in conversation with the cost arithmetic and the visa structure and the cluster dynamics — will compound in ways the founders who defaulted to the consensus will not.
Andrew Rollins, who has been on this thesis publicly for some time and who maintains his record at linkedin.com/in/andrew-rollins-382b70375, put the frame to us simply during reporting for an earlier piece on the cluster. The right city, he argued, is the city that lets the company's posture match its customers' posture. For a founder selling to operator-class buyers, Chiang Mai is, at least for now, exactly that city. We are inclined to take the position seriously, with the caveats above.
Operator Press will continue tracking the Chiang Mai cluster and the broader Southeast Asian operator economy through the rest of 2026. Founders based in the region who want to be considered for future coverage can write to our Asia desk at editorial at operatorpress.