https://github.com/sujalpatel92/graph-rag-neo4j-starter
Starter for evidence-grounded QA with GraphRAG on Neo4j and FastAPI. Ingest docs, build a graph, query with citations.
https://github.com/sujalpatel92/graph-rag-neo4j-starter
docker fastapi generative-ai graph-rag llm mlops neo4j nlp python streamlit
Last synced: 20 days ago
JSON representation
Starter for evidence-grounded QA with GraphRAG on Neo4j and FastAPI. Ingest docs, build a graph, query with citations.
- Host: GitHub
- URL: https://github.com/sujalpatel92/graph-rag-neo4j-starter
- Owner: sujalpatel92
- License: mit
- Created: 2025-09-16T20:43:49.000Z (22 days ago)
- Default Branch: main
- Last Pushed: 2025-09-16T21:24:31.000Z (22 days ago)
- Last Synced: 2025-09-16T23:52:39.751Z (21 days ago)
- Topics: docker, fastapi, generative-ai, graph-rag, llm, mlops, neo4j, nlp, python, streamlit
- Homepage:
- Size: 8.79 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# graph-rag-neo4j-starter
> Starter template for evidence-grounded question answering using GraphRAG with Neo4j and FastAPI.
## Features
- Local-first Docker Compose stack with Neo4j, OpenSearch, Redis
- Ingestion for PDFs, HTML, and text files with provenance
- Query modes: text, entity, relationship, combined with citations
## Quickstart
```bash
# clone
git clone https://github.com/{{GITHUB_USER}}/{{REPO_SLUG}}.git && cd {{REPO_SLUG}}
# Python setup with uv
curl -LsSf https://astral.sh/uv/install.sh | sh && uv venv .venv && uv pip install -e ".[dev]"
# run example service
uv run python -m {{PACKAGE_ENTRYPOINT}}
```
## Architecture (Sketch)
- Ingestion → Extraction → Graph writer → Indexer → Query API → UI
- Include Mermaid or PNG diagram later.
## Configuration
- Copy `.env.example` to `.env` and fill values.
## Development
- Lint: `uv run ruff check .` | Format: `uv run black .`
- Test: `uv run pytest -q`
## Roadmap
- MVP → Observability → Deploy
## License
MIT