Adrian Li

about

I am a data engineering and AI/ML practitioner with experience building scalable ETL pipelines, cloud data workflows, machine learning models, and agentic RAG applications. I have designed Databricks pipelines processing more than 5 million records daily, migrated legacy workflows to event-driven orchestration, and built reusable CI/CD frameworks with Jenkins and Git. My work spans SQL and Python data processing, recommendation and propensity modeling with XGBoost, sentiment analysis, and production-oriented GenAI systems. I build LangGraph and LangChain agents that combine structured data retrieval, vector search, web search, tool routing, reflection loops, source citations, and evaluation frameworks. I also develop APIs and interfaces using FastAPI and Streamlit, with experience in PostgreSQL, pgvector, Docker, LangSmith, and Pytest. I can help with data pipeline development, workflow automation, ML modeling, RAG systems, AI agent orchestration, data retrieval architecture, and evaluation of AI application quality, latency, and groundedness.