Our Story

Teaching AI development the long way round

Bhumi AI was built around a single belief — that rushing through technical material is the most reliable way to not understand it.

← Back to Home

About Bhumi AI

Where the name comes from

Bhumi — from Pali and Sanskrit — means ground, earth, foundation. It was chosen deliberately. The school started from the observation that most technical education in AI rushes students toward frameworks and tools before they've developed any feel for the terrain underneath. Bhumi AI was set up to work the other way around.

The school opened in Bangkok in 2022, initially running small cohort sessions in mathematics and probability for people moving into data roles. The response was consistent: students said they'd finally understood things they'd tried to learn quickly before, and hadn't retained. From that, the curriculum grew — cautiously — into MLOps and eventually into the current AI Research Methods programme.

We're based in Chatuchak, a neighbourhood that mixes old market culture with newer technology offices. It's a useful place to run a school that sits between the traditional and the applied.

Mission

What we're trying to do

The AI field moves quickly. We don't think the answer to that is to teach people to move quickly too. Our aim is to give students a stable enough understanding of the fundamentals that when the tooling changes — and it will — they can reason about the new situation rather than starting over.

This means we spend time on things that don't make obvious headlines: careful mathematical reasoning, reading research papers with patience, understanding what production systems actually require. These are slower to teach. They tend to stay learned.

We keep cohorts small so that each student gets meaningful contact with instructors. We don't believe online education at scale delivers the same experience as working in a smaller, more attentive setting.

The Team

People behind the courses

Nattapong Promchote

Lead Instructor — Mathematics & ML

Spent eight years working in quantitative modelling before moving into education. Teaches the mathematics foundations course and oversees the MLOps curriculum design.

Siriya Klangthong

Research Methods Instructor

Background in cognitive science research. Leads the AI Research Methods programme, with particular focus on experiment design and the ethics of empirical work in AI.

Anya Witthaya

MLOps & Systems Instructor

Has run machine learning infrastructure at two Bangkok-based companies. Brings direct production experience to the MLOps course, with a focus on practical tooling and realistic deployment scenarios.

Our Standards

How we hold ourselves to account

Curriculum Review

Course content is reviewed before each new cohort. Instructors flag what students found unclear, what took longer than expected, and what produced better outcomes. Changes are made on that basis.

Cohort Size Limits

We cap enrolment to preserve instructor contact quality. This is not a growth strategy — it's a teaching decision. When cohorts are too large, individual student progress becomes harder to track.

Student Data Privacy

We collect only what's needed to administer courses and communicate with students. Student work is not used commercially. Data is retained only for as long as necessary.

Ethics in the Curriculum

Research ethics, bias in datasets, and responsible deployment are not elective topics in our programmes. They are part of the core material — because they're part of the actual work.

Feedback at Every Stage

Students receive written feedback on assignments — not just scores. For the Research Methods programme, structured mentor review is built into the timetable. We treat feedback as part of the teaching.

Compliance with Thai Educational Standards

Bhumi AI operates in accordance with relevant Thai regulations governing private educational institutions and data protection law, including the Personal Data Protection Act B.E. 2562.

Our Approach in Detail

Applied AI education that starts from the ground up

Bhumi AI teaches applied AI development through three connected programmes: a mathematics foundations course, a practical MLOps course, and a long-form AI research methods programme. Each is designed to build on a clear understanding of fundamentals, rather than assuming that familiarity with tools is the same as competence.

The school is physically located at 145/9 Phaholyothin Road in Chatuchak, Bangkok — a working studio space where small groups meet for collaborative sessions. Online participation is possible for learners elsewhere in Thailand, with sessions structured to maintain the quality of instructor interaction.

The mathematics programme addresses one of the most common barriers to deeper ML work: gaps in the mathematical reasoning that underpins standard algorithms. By working through linear algebra, calculus, and probability in a structured way — with weekly problem sets rather than lecture-only delivery — students leave with a working relationship to the material rather than a surface acquaintance.

The MLOps course is built around the recognition that most ML education stops at model training. What makes production systems work — versioning, monitoring, deployment pipelines, handling data drift — is a distinct body of knowledge that requires direct engagement with real tooling.

The AI Research Methods programme exists because reading papers, designing experiments, and engaging with the peer-reviewed literature are learnable skills — and because an increasing number of practitioners want to develop them. The programme is deliberately long, because that work takes time to develop well.

Curious about a particular course?

Drop us a message and we'll tell you what background suits each programme and what you might expect from the experience.

Get in Touch