Three kinds of user
The self-directed learner
You are working through a hard field on your own — machine learning, quantitative finance, cryptography, a language — and you have collected far more material than you have actually absorbed. You do not need another summary. You need something that decides what you should study next, and then refuses to let you believe you have learned it until you can show it.
What you get: your own material turned into a knowledge map, a tutor that argues with you inside it, and a review schedule that keeps what you have earned.
The teacher or lecturer
You have a syllabus, notes and slides that took years to get right, and you would like students to be tutored on that — your definitions, your notation, your scope — rather than on whatever a general chatbot improvises. You also want to see where a cohort is actually stuck, before the exam tells you.
What you get: upload or sync your material, approve the extracted knowledge components and their prerequisites, run a course with a roster and per-student progress reports, and see which topics students keep asking for.
The institution
You want a tutor bound to your own courses, with each student seeing only what they are entitled to see, and with that boundary enforced somewhere more serious than the front end.
What you get: institutions, courses, rosters and revocable entitlements, with isolation enforced by database row-level security.
Where it stands: the predecessor was piloted for a full academic year in 2024–2025 — at Corvinus University of Budapest, in Solana Superteam's educational tracks, and in the founder's own teaching at Óbuda University. The rebuilt platform took its first institution through end to end in August 2026. Setting up a new one is still done with an operator's help rather than self-service.
What is being learned on it
Seventeen subjects are wired into the content pipeline, most of them technical and demanding — machine learning, neural networks, transformers, AI agents, retrieval-augmented generation, MLOps, quantitative principles, algorithmic trading, crypto and forex material, secret management — alongside English-language study and a Hungarian-language AI fundamentals course.
That mix is not accidental. The system was built by someone using it to learn algorithmic trading and English at the same time, which is also why teaching in your first language while your English catches up is a first-class feature rather than an afterthought.