🧠 infuse-ai

infuse-ai is a CLI-based Retrieval-Augmented Generation (RAG) system that intelligently scans PDF documents and uses a Large Language Model (LLM) to provide highly relevant, source-backed answers.

📚 Ask questions about your documents. Get accurate, contextual, and cited responses — straight from your terminal.

---

✨ Features

• 🔍 PDF-Based Retrieval: Extracts and indexes content from PDFs
• 🧠 LLM-Powered Answers: Uses a large language model to generate contextual responses
• 🧾 Source-Backed Results: Each response is linked to the relevant source content
• 🛠️ CLI Tool: Simple command-line interface for querying your data
• 🚀 Modular Architecture: Built with Node.js for extensibility and future integrations

---

📦 Tech Stack

• Node.js – Core implementation
• AWS Bedrock (Claude) – LLM backend
• PDF Parser – To extract textual content from documents
• Vector Search – Embedding and similarity matching (planned)

---

📁 Supported Input

• PDF documents (more types coming soon: .txt, .csv, .docx, and web links)

---

💡 Use Case

• Instantly query research papers, legal documents, manuals, or internal reports
• Extract meaningful insights with traceable sources
• Ideal for developers, researchers, and knowledge workers

---

🚀 Getting Started

Clone the Reposiory

git clone https://github.com/hemupadhyay26/infuse-ai.git
cd infuse-ai
yarn install
docker compose up -d
cp .env.example .env

paste the needed credential to .env file and start server

yarn db:migrate
yarn dev