Portfolio

The Chunker (GitHub)

Standalone chunking engine that turns code and text files into semantic, token-aware chunks for RAG and LLM pipelines.

The Embedder (GitHub)

End-to-end Retrieval-Augmented Generation stack that handles ingestion, chunking, indexing, and serving results through Open WebUI. This monorepo packages every moving part: a production-grade ingestion CLI, a reusable chunker library, a FastAPI connector for Qdrant, and local/Docker/AWS deployment assets.