Mnemo

Mnemo is a lightweight, self-hostable memory middleware REST API for AI agents. It provides:

• Episodic memory: immutable raw turns stored in SQLite
• Semantic memory: bi-temporal triplet edges stored in SQLite + Qdrant vectors
• User profile: persistent key/value facts stored in SQLite + cached in Redis

Quickstart

1. Create a virtualenv and install dependencies (using uv recommended):

uv venv
source .venv/bin/activate
uv pip install -e ".[dev]"

2. Copy env file and set keys:

cp .env.example .env

3. Start Redis (Docker recommended):

docker compose up -d redis

4. Run the API:

uvicorn mnemo.app.main:app --reload

5. Open docs:

• http://localhost:8000/docs

Notes

• Qdrant runs in embedded local mode by default (no Docker required).
• Extraction runs asynchronously via ARQ worker (1-turn lag by design).