What chat.eduba.io does
Four core capabilities, built around how people actually learn and work with AI agents. Each one shows up inside a discussion, where the system prompt, participants, and history are all in one place.
Navigate hierarchical knowledge
Explore turns a topic into a structured tree you can move through with the agent. Drill into a branch when it matters, fold it back up when it doesn't, and keep the trail of how you got there.
- AI-guided exploration scoped to the current discussion.
- Hierarchical structure so depth is intentional, not accidental.
- Drill into any branch on demand and keep the trail of how you got there.
Practice, not just consumption
Card actions are short, game-based interactions inside a discussion. They turn passive reading into deliberate practice, with quick feedback loops that build skill instead of just exposure.
- Every card is generated based on your discussion content.
- Quick interactions for daily reinforcement.
- Templates available when you create a new discussion.
Review what you actually covered
Flashcards are generated from your own discussions, not a generic deck. They keep the things you worked through retrievable later, so a session you had once becomes knowledge you keep.
- AI-generated cards tied to your discussion history.
- Aligned to your goals and the system prompt you set.
- Useful for review, retention, and prep.
The system underneath
Every discussion has a topic, a system prompt, and a set of participants. Participants can be people or agents. That structure is the same whether you're using chat.eduba.io as an individual or running it inside an organization.
- Discussions with topic, system prompt, and full message history.
- User personas and agent personas as first-class participants.
- Markdown documents attached to discussions for shared context.
- Agent personas built on top of admin-approved providers.