These are accounts from learners across our three courses — written in their own words, reflecting what the experience was actually like.
Back to HomeI had used statistical methods in my job for years without understanding the assumptions behind them. The first few weeks of the statistics course felt uncomfortably foundational — but by week four, I realised the foundation was exactly what I had been missing. My mentor never made me feel behind for asking basic questions.
The language models course is the first time I felt I properly understood what a language model is actually doing — rather than just how to write prompts that happen to work. The ethical sections were genuinely useful, not token gestures. My one wish is that there were slightly more worked examples in the later weeks.
I enrolled in the mentorship track after completing both taught courses. Having a mentor dedicated to my project — a classification system for Thai medical records — made the difference between something I would have abandoned and something I am proud of. The sessions were scheduled around my job, which was essential.
Taking this fully online while working in Khon Kaen was straightforward. The pacing was genuinely manageable — I studied on evenings and one weekend morning per week. The written feedback from the mentor was detailed enough that I rarely needed to ask follow-up questions, though they responded promptly when I did.
I appreciated that the language models course did not treat prompting as the whole story. The architecture sections required more patience than I expected but paid off when I started debugging why a model was producing inconsistent outputs. I went from confused to reasonably competent over nine weeks — at my own pace.
I came into the mentorship track without a clear project idea, which I worried about. Supaporn — my mentor — helped me scope something realistic within my first session. We ended up building a customer segmentation model for the retail dataset my employer uses. I have since presented it internally. That was not something I expected to happen.
Surachai had been running A/B tests at work for three years but had never been confident in how to interpret the p-values his tools produced. When results were ambiguous, he had no framework for making a principled call.
He studied the statistics course over eight weeks, spending extra time on the hypothesis testing module. His mentor worked through two of his workplace datasets with him as exercises — grounding the concepts in familiar material.
Surachai now leads the statistical review for his team's experiments. He has documented a testing framework for colleagues that draws directly on what he covered in the course. He describes it as the most practically useful thing he has studied in five years.
Wiraphon wanted to build a document classification system for Thai medical text but had no experience applying NLP to a low-resource language. He had read widely but had never put anything together end to end.
Over four months he built a classification pipeline from data collection through to evaluation. His mentor helped him scope the project sensibly in the first session, then reviewed his code and approach at each fortnightly check-in.
The project reached 84% classification accuracy on held-out data — a result Wiraphon considers defensible given the dataset size. He has written it up and shared it with a hospital research unit as a potential pilot collaboration.
Parichat's employer had years of customer transaction data but no structured approach to segmentation. Manual groupings were based on intuition rather than pattern, and she had been asked to develop something more reliable.
Working with her mentor, she explored clustering approaches and settled on a method suited to the data structure. The mentor reviewed her feature selection reasoning carefully and pushed back on one assumption that would have introduced bias.
The segmentation model identified five distinct customer groups not visible in the previous manual approach. Parichat presented her methodology to the company's marketing leadership, and the model is now informing campaign planning for three product lines.
We are happy to answer them directly — no pressure to commit first.
18 Sathorn Road,
Bangkok 10120
Mon–Fri: 9:00–18:00
Sat: 10:00–15:00
Member since 2022. Active contributor to community discussions on pedagogical design in technical education.
Verified compliant with Thailand's Personal Data Protection Act. Annual review conducted May 2025.
Featured school profile published in the community newsletter, April 2025. Recognised for mentorship quality.
Tell us where you are in your learning, and we will suggest where to start. No obligation — just a considered response from someone who knows the courses well.
Get in Touch