A goal as input
Generation starts with the learning goal and produces sections and content for the path.
An educational prototype with AI-generated learning paths, study texts, flashcards and minigames, built as a team at ICP WCHL.
The team explored how to turn a learning goal into a study path. Kaisen uses that goal to generate sections and content, then combines path context with the user's question when asking the AI for help.
I implemented frontend and backend features for registration, login, profiles and learning-path CRUD. I integrated Gemini for goal-based path generation, contextual answers and specialized minigame calls: each call received path and minigame context and returned JSON.
The prototype had working study texts and flashcards and won first place at ICP WCHL 2025. It was a team hackathon delivery with implemented features and parts of the scope still incomplete.
Community features, feed recommendations and ranking were not implemented. Handling of model failures and invalid responses was limited. This case describes a prototype rather than a complete production product.
Generation starts with the learning goal and produces sections and content for the path.
Questions sent to the model combine the user's message with the learning-path context.
The team chose Gemini considering free-tier limits. Specialized minigame calls returned JSON data.
contact.md
For software engineering opportunities and internships, with room to contribute through applied AI and data. I reply within 3 days.