Modelcraft learning environment

Learning to Build AI Should Feel Like Building Something

Modelcraft was set up to make that possible — with clear tracks, real projects, and room to ask questions at every step.

← Back to Home

Built from a Shared Frustration

Modelcraft started as a conversation between two developers who had both spent time trying to learn AI through online platforms — and found them either too shallow or too dense to be useful in practice.

The workshops and video libraries were fine for getting an overview. But when it came time to actually write code, prepare data, and make something that worked, there were gaps that no amount of re-watching solved. The missing piece was structure, feedback, and a place to get stuck without feeling like you were falling behind.

In 2022, we began running small, in-person sessions for a handful of learners in Phuket. Each session was built around a specific task — not a lecture, not a slideshow, but a working exercise with a real output. When the sessions filled up and we started getting requests from outside Thailand, we built the online track structure you see now.

The name Modelcraft is deliberate. A model, in machine learning, is something you train. A craft is something you develop over time, with practice. We think those two ideas belong together.

Our Mission

To give learners a clear, structured way to build AI skills — through real projects, honest feedback, and a pace that fits around actual lives.

Our Vision

A school where someone with no coding background can arrive, work steadily through a track, and leave with a project they built themselves — and the skills to keep going.

Our Commitment

We do not make promises about outcomes. We make commitments about process — honest materials, available mentors, and feedback that is specific enough to be useful.

People Behind the Workbench

DK

David Kramer

Lead Developer & Curriculum Designer

David has been building machine learning tools for over eight years across Southeast Asia and Europe. He designs all track content and runs the Capstone mentorship sessions.

NP

Naphat Pongpat

Computer Vision Specialist & Track Mentor

Naphat holds a background in image processing and has worked on applied vision projects in agriculture and logistics. He leads the Computer Vision Project Track and reviews learner assignments.

SW

Sara Wichit

Learner Experience & Community Lead

Sara manages the learner community, coordinates cohort scheduling, and makes sure questions in the forum get answered. She also runs the onboarding process for new enrolments.

Standards We Hold Ourselves To

Materials That Work in Practice

Every exercise is tested by someone who did not write it. If an instruction is confusing or a code block does not run cleanly, it gets rewritten before it goes into a track.

Data Privacy by Default

Learner data is stored securely and is not shared with third parties for advertising or analysis. We collect only what is needed to run the programme.

Regular Content Review

AI tooling moves fast. We review track materials on a rolling basis and flag when something has changed significantly — learners are notified when updates affect their track.

Accessible Language

Technical terms are explained when they first appear. No assumed background means materials are written for someone starting fresh, not for someone already familiar with the field.

Timely Mentor Responses

Questions posted in the forum or flagged to a mentor receive a response within one working day. One-to-one sessions are scheduled within 48 hours of a request during the track.

Honest About Limitations

We do not overstate what these tracks will do for learners. Skill development takes time and continued practice beyond any course. We say that clearly, and we try to support the continuation.

AI Development as a Practical Discipline

Modelcraft is an online school focused on AI development — specifically, on the skills involved in building, training, and evaluating models. The school is based in Phuket, Thailand, and runs tracks that learners access remotely or, where schedules allow, in person.

The three tracks on offer cover different stages of the learning path. The Programming for AI workshop is designed for people starting from scratch with Python — it focuses on writing functional code, working with data files, and understanding what code is doing rather than just copying it. The Computer Vision Project Track builds on those foundations and moves into the specific domain of image-based models, which are widely used in fields from logistics to environmental monitoring. The Capstone Build & Mentorship track is for learners who want to take on a larger, self-directed project with structured support.

The school keeps cohort sizes small. This is not a marketing decision — it is a practical one. Feedback on assignments takes time to write properly, and one-to-one mentorship sessions take time to run well. Keeping numbers manageable makes both of those things possible.

Modelcraft is not affiliated with any specific tooling company or platform. The tracks use open-source libraries and standard tools that learners can continue using after the programme ends. The aim is to develop skills that transfer — to future projects, to further study, or to working contexts where AI knowledge is increasingly useful.

Find the Track That Fits Your Starting Point

Get in touch to ask which track makes sense for where you are. We will talk through the options and give you a clear picture of what to expect.

Get in Touch