DocWeave

DocWeave is an enterprise-ready, agentic Retrieval-Augmented Generation (RAG) platform for document-heavy workflows, with a strong fit for regulated domains.

It helps teams ingest heterogeneous documents, retrieve grounded evidence, and deliver auditable answers through APIs and UI surfaces.

Why DocWeave

• End-to-end RAG workflow from ingestion to answer generation
• Multi-format document support: PDF, CSV/XLSX, JSON/JSONL, TXT/MD
• Hybrid retrieval (vector + BM25) with optional reranking
• Agentic orchestration for retrieval, grounding, and analysis
• Multi-tenant patterns for isolation and scale
• Production-oriented API, worker, observability, and runbook support

Core Capabilities

• Ingestion pipeline: parse -> normalize -> chunk -> embed -> index
• Vector backends: pgvector, qdrant, chroma
• LangGraph agent flow: route -> retrieve -> ground -> analyze -> answer
• FastAPI application layer for serving and integration
• Celery worker for asynchronous/background processing
• Security and operations guides for deployment hardening

Architecture and Docs

• Quickstart: docs/QUICKSTART.md
• Architecture: docs/ARCHITECTURE.md
• API: docs/API.md
• Security: docs/SECURITY.md
• Runbooks: docs/RUNBOOKS.md

Quick Start

make dev
make run

In a separate terminal:

make worker

Optional sample indexing:

make index

Local Development

Available Make targets:

• make dev - create virtual environment and install editable dependencies
• make run / make api - start the FastAPI app with reload
• make worker - start Celery worker queues
• make index - build sample index from data/samples
• make test - run test suite
• make lint - run Ruff lint checks
• make type - run MyPy type checks
• make docker - run with Docker Compose

Deployment Notes

DocWeave includes deployment assets under deploy/ and docker/ for containerized and Kubernetes/Helm-based environments.

Before production rollout:

• Configure secrets and environment variables securely
• Validate tenant isolation and data boundaries
• Enable tracing/metrics and log redaction policies
• Review docs/SECURITY.md and docs/RUNBOOKS.md

Project Scope

This repository focuses on practical, production-minded RAG patterns with tutorial-level clarity for fast onboarding.

License

MIT - see LICENSE.