At Databricks, I design, build, and operate production LLM applications and RAG services using Mosaic AI, Vector Search, and Amazon Bedrock. This work improved response accuracy by 20%.
I own services end to end, from architecture and implementation through deployment, evaluation, and monitoring, achieving 95% SLA compliance. I also build Python and FastAPI APIs that connect LLM applications with enterprise data and tools.
At Databricks, I automated data and ML pipelines with Databricks Workflows, Auto Loader, Delta Live Tables, PySpark, and Delta Lake, reducing data processing time by 30%. I evaluate generative systems with MLflow and tune chunking, embeddings, and semantic retrieval before releases.
Previously, at Deloitte, I developed classification and predictive ML models with Python, Scikit-learn, TensorFlow, and Keras, improving predictive analytics performance and customer experience by 25%. I also worked on AWS ETL and streaming pipelines and built secure REST APIs.
