Cognitive Companion

On-premise AI system for senior care: safety monitoring, activity tracking, and cognitive engagement. All inference runs locally.

The backend is a FastAPI BFF gateway for the continuous tracking, scene analysis, person identification, semantic memory, and TTS services. A Vue 3 admin UI provides camera management, live tracking, privacy zone configuration, rule authoring, and dementia signal dashboards. A senior-facing voice companion runs through the PWA.

Documentation: silvermind-project.github.io

Architecture

flowchart LR
    Edge[Edge Devices] --> CC[Cognitive Companion]
    CC --> AI[AI Services]
    CC --> CTS[Continuous Tracking]
    CTS --> CC

Loading

The system ingests camera frames and sensor events, evaluates them against caregiver-authored rules with composable per-rule pipelines, and dispatches notifications across seven output channels.

Quick start

cp .env.example .env
docker compose -f docker-compose.db.yml -p nanai up -d  # shared Postgres
docker compose up -d
make init-db

See the documentation site for person enrollment, pipeline configuration, and operations guides.

Development

# Backend
uv run --project backend uvicorn backend.main:app --host 0.0.0.0 --port 8000 --reload

# Frontend
cd frontend && npm install && npm run dev

# Quality gates
make check              # lint + strict mypy on core + core tests
make check-all          # adds services tests

License

AGPL-3.0-or-later