We Built Bhutan's First AI Company From a College Dorm. Here Is What That Actually Looked Like.

There is a version of this story that sounds inspiring from the outside. Students from a small Himalayan kingdom, barely out of college, building AI systems for banks and airlines. Bloomberg wrote about it. Rest of World covered it.

The version from the inside looked different. It looked like a lot of late nights, broken prototypes, clients who were skeptical, and a country with almost no AI ecosystem to learn from or lean on.

This is that version.

Why We Started

Bhutan's institutions were running on systems that had not meaningfully changed in decades. Banks were handling thousands of customer queries manually. Airlines were answering the same 20 questions on repeat. Government agencies were processing paperwork that could be automated in weeks.

We were students at Gyalpozhing College of Information Technology, watching this happen, and feeling an impatience that is hard to explain. The tools existed. The problems were obvious. Nobody was building the solutions.

So we did.

We did not wait for someone from outside Bhutan to come fix it. We just started building.

The First Real Deployment

Our first serious client was Bhutan National Bank. The brief was simple: build a conversational AI assistant for their website. The reality was harder. BNB ran on legacy SOAP and XML systems. There was no clean API. Integration meant building custom middleware from scratch to connect a modern AI layer to 20-year-old infrastructure.

We built it. It went live. Users started using it.

That first deployment taught us more than two years of coursework. You learn differently when the system is in production and real people are depending on it.

What Came After

Bhutan Development Bank came next. Then Drukair. Then more institutions. Each deployment was harder than the last, and each one made us faster and more capable.

By 2024, we had 100,980 users across 9 countries interacting with systems we had built. Not pilots. Not demos. Production systems running every day.

Bloomberg noticed. Rest of World noticed. We appreciated the coverage. But what mattered more was that the systems kept working.

Where We Are Going

Customer support was where we started. It is not where we are going.

The real opportunity, the one we are now building toward, is deeper. Enterprise and government institutions across South Asia need AI systems that run their operations, not just their helpdesks. Custom software, internal automation, digital transformation at scale.

That is what NoMind is becoming.

We are always open to serious conversations. /contact