An AI-powered assistant that reads documents from a folder and answers questions, gives recommendations, and generates templates/playbooks — using Retrieval-Augmented Generation (RAG).
This AI agent can:
- Connect to a folder and ingest documents
- Use embeddings to understand document content
- Answer questions and generate recommendations based on the files
- Create structured outputs like templates, checklists, and playbooks
- Python 🐍
- LLMs
- LangChain for agent orchestration
- Azure AD + Microsoft Graph API for SharePoint access
- Vector DB for retrieval
git clone https://github.com/teofizzy/doc-savvy.git
cd doc-savvy