What a slower, more careful approach to AI education makes possible

Pranee Lertkul, who leads programme design, spent fifteen years as a mathematics lecturer before building Lotus Studio's curriculum. That background shapes how material is sequenced and explained — the pacing, the choice of examples, the decision about which abstractions to introduce early and which to defer. These choices are not arbitrary; they come from watching many learners work through the same difficulties over many years.

Thanapat Suwannarat, who leads the generative AI programme, brings applied research experience alongside his teaching work. He is familiar with the literature at a working level, which means he can explain not just what a paper says but what choices the authors were navigating when they wrote it.

• Curriculum developed by working specialists
• Teaching experience spanning 15+ years
• Content informed by ongoing engagement with the field

The Generative AI and Diffusion Models programme is built around methods that are active areas of development in the field. The papers and frameworks covered were selected because they represent how practitioners actually work with these systems today, not because they are well-established enough to feel safe to include.

Programmes are reviewed twice a year to catch developments that have shifted the landscape meaningfully. When something changes in a way that affects what a working understanding of the field requires, the curriculum is updated. We describe when each programme was last revised so learners know what they are getting.

• Biannual curriculum review cycle
• Exercises using current open-source tooling
• Content sourced from active research, not aged summaries

Questions and difficulties are handled by the instructors responsible for the programme, not routed through a general support inbox. This matters because the kind of question that arises when you are working through a proof or debugging an implementation often requires contextual understanding that only the course designer actually has.

The AI Research Pathway includes scheduled one-to-one mentorship sessions with Nattida Wongchai, arranged at a time that suits the learner. Those sessions are substantive — focused on the learner's specific research questions and the particular challenges they have encountered.

Lotus Studio charges ฿2,700 for Mathematics for AI, ฿5,600 for Generative AI and Diffusion Models, and ฿8,200 for the AI Research Pathway. These figures include everything described in the programme — materials, exercises, and where applicable, mentorship. There are no additional fees to access resources, submit work for feedback, or receive support during the programme.

We do not offer tiered access where a lower price buys a diminished experience. If a learner enrols in a programme, they have full access to what it contains from the start.

Learners who complete Mathematics for AI can work through the mathematical sections of ML papers without needing to look up every other term. They have a stable foundation to bring to more advanced study. Learners who complete the Generative AI programme can read and critically engage with new papers in the diffusion model space, and implement core components from scratch using established frameworks.

Learners on the Research Pathway develop an original research contribution — small in scope, but genuinely their own — and the communication practices needed to present it. Several have gone on to submit to workshops at regional and international AI conferences.