graphaura

Bring your memories back to life with GraphAura. Transform photos and docs into an AI-powered 3D knowledge graph of people, places, and moments to explore your memories and share insights.

Repositories

• graphaura: this monorepo with frontend and backend
• R2R: external service for ingestion, chunking, embeddings, and RAG

Tech stack

• Frontend: Next.js 15, React 19, Tailwind CSS, react-force-graph-3d
• Backend: FastAPI (Python 3.13), structlog
• Data: Neo4j, PostgreSQL + pgvector, Redis
• AI: R2R for ingestion, NER, embeddings, and RAG

Architecture

graph TB
  U[Browser] --> FE[Next.js Frontend]
  FE -->|REST| BE[FastAPI Backend]
  BE -->|HTTP| R2R[R2R Service]
  BE -->|bolt| NEO[(Neo4j)]
  BE -->|asyncpg + pgvector| PG[(Postgres)]
  BE -->|Redis| REDIS[(Redis)]
  R2R -->|entities + embeddings| BE

Loading

Data flow

flowchart LR
  Files[Files, Photos, Notes, PDFs] --> Upload[/Upload via UI or API/]
  Upload --> R2RIngest[R2R ingest + chunk + embed + extract]
  R2RIngest --> Entities[Entities + Relationships]
  Entities --> NEO[(Neo4j)]
  R2RIngest --> Embeds[Embeddings]
  Embeds --> PG[(Postgres pgvector)]
  Query[User query] --> API[Search APIs]
  API --> Hybrid[Hybrid search]
  Hybrid --> Docs[R2R document search]
  Hybrid --> Graph[Graph traversal]
  Graph --> FEVis[3D graph UI]
  Docs --> FEVis
  NEO --> API
  PG --> API

Loading

Monorepo layout

graphaura/
  backend/        FastAPI app, services, and APIs
  frontend/       Next.js app with 3D graph visualization

Quick start (dev)

1. Dependencies

   • Neo4j 5 running locally
   • Postgres with pgvector extension
   • Redis
   • R2R service available at http://localhost:7272

2. Backend

   • cd backend
   • uv sync
   • copy .env.example to .env and set connection strings
   • uv run uvicorn src.main:app --reload

3. Frontend

   • cd frontend
   • npm install
   • npm run dev

Key endpoints

• GET /health - service health
• GET /metrics - vector and graph counts
• POST /api/v1/documents/upload - upload and process document via R2R
• GET /api/v1/documents/{id} - document metadata
• POST /api/v1/documents/create-graph - build graph from documents
• POST /api/v1/graph/entities - create entity
• GET /api/v1/graph/entities/{id} - fetch entity and relationships
• POST /api/v1/search/hybrid - hybrid search across documents and graph

Configuration

Set via environment variables (see backend/.env.example and backend/src/config.py). Common keys:

• POSTGRES_* - Postgres connection and database name
• NEO4J_* - Neo4j connection
• R2R_BASE_URL - R2R base URL
• REDIS_* - Redis connection
• VECTOR_DIMENSION, SIMILARITY_THRESHOLD, MAX_GRAPH_DEPTH

Notes

• See backend/README.md for detailed APIs and development commands.
• Frontend runs on http://localhost:3000. Backend runs on http://localhost:8000 with OpenAPI docs at /docs.