projects/

Browser-based RAG learning app that runs 100% locally — no server, no API keys, no data leaving the tab.

• - Runs embeddings and generation entirely in-browser via Hugging Face Transformers.js.
• - Visualizes the full pipeline: chunking, embeddings, retrieval, and generation in real time.
• - Upload your own text, ask questions, and watch every retrieval step as it happens.

Retrieval-Augmented Generation system with multi-agent workflows for context-aware answers over private documents.

• - Multi-agent orchestration with LangGraph for retrieval, grading, and re-query loops.
• - Advanced document retrieval over a vector store with relevance filtering.
• - Context-aware responses grounded in retrieved source material.