Machine Learning Engineer
CurrentBuild retrieval augmented generation (RAG) system from scratch to deliver LLM reasoning shopping experiences "Rufus", which has been launched globally in Amazon Mobile/Web apps.My recent focus around LLMs is on building blocks to simplify model development and release (SDK, training, evaluation, inference, RAG, batching, ...) and designing distributed systems to deal with simple and complex workflow orchestration and peak traffic, as well as delivering automation tools that impact our daily work.