Finrag
Provides secure, role-based access to query company documents using Retrieval-Augmented Generation.
About
FinRAG is a sophisticated Role-Based Access Control (RBAC) Retrieval-Augmented Generation (RAG) system designed to enable secure and trusted querying of internal company documents. It ensures that employees, managers, and C-level executives only access information relevant to their roles, providing secure answers. By enriching documents with metadata and citations and integrating with Claude Desktop's Model Context Protocol, it delivers reliable responses with transparent sourcing, making it a flexible and practical solution for handling various document types from engineering reports to financial summaries.
Key Features
- 0 GitHub stars
- Optional Streamlit UI and FastAPI gateway for flexible deployment
- Seamless integration with Claude Desktop (MCP tools)
- Retrieval-Augmented Generation (RAG) with source citations
- Local Qdrant for efficient vector search
- Role-Based Access Control (RBAC) for secure information access
Use Cases
- Accessing engineering documentation and service lists tailored to engineering roles
- Securely querying HR policies and employee handbooks based on role
- Summarizing financial reports and revenue drivers for finance teams