Satyaaakash/Technical-Documentation-Assistant

Technical Documentation Assistant

This project demonstrates an AI-powered assistant for technical documentation using retrieval-augmented generation principles.

Features

• —Vector Search: Find documentation based on semantic similarity
• —Multiple Source Integration: Combines information from PyTorch, TensorFlow, and Hugging Face Transformers
• —Web Search Fallback: When information isn't found in the knowledge base, falls back to web search

How to Use

Simply ask a technical question in the input box and get relevant information from the knowledge base or web search.

Sample Questions

Try some of these questions:

• —What is PyTorch?
• —How does TensorFlow compare to PyTorch?
• —What are transformers in machine learning?
• —How to implement a neural network in PyTorch?
• —What is the difference between RNN and Transformer models?

Technologies Used

• —Vector Embeddings (BAAI/bge-large-en-v1.5)
• —Qdrant Vector Database
• —Streamlit for UI