Isaiah Abhulimen

I've built a production-oriented RAG Knowledge Assistant for grounded question answering across multiple books, using Python, FastAPI, ChromaDB, BM25, Sentence Transformers, cross-encoder reranking, Docker, and Railway.

I designed its hybrid retrieval pipeline, document ingestion, semantic chunking, incremental indexing, persistent vector storage, background processing, resumable checkpoints, question routing, decomposition, and benchmark-based evaluation.

I also added API authentication, rate limiting, structured logging, health checks, error handling, automated testing, and CI/CD with GitHub Actions and GitHub Container Registry.

Alongside my AI/RAG work, I've spent 8+ years founding and operating Ero Kings Arts & Signages, leading client delivery, procurement, operations, and teams of 5–10 workers while building my career in AI/RAG engineering and Python backend development.