Use case / Open Source AI Memory Database

Education Tutor Agent Memory

Tutor agents need long-term learner memory so instruction can adapt to goals, misconceptions, progress, and preferred explanation style over time.

Read the docs Star on GitHub

The problem this solves

Without durable learner memory, tutoring agents repeat the same diagnostic questions and fail to build on prior progress. Without lifecycle controls, old misconceptions can keep polluting future lessons.

How Memorose handles it

Memorose can store learner-scoped events, retrieve relevant educational context, and update or forget stale learning state as the student improves.

runtime example

curl -X POST "$BASE_URL/v1/users/student-19/streams/algebra-tutor/events" \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "agent_id": "tutor-agent",
    "content": "Student understands linear equations but confuses slope with intercept."
  }'