At Upstart, I lead development of large-scale machine learning, deep learning, and LLM-based ranking and recommendation systems. I architect end-to-end ML pipelines and build transformer models for ranking, classification, and sequence prediction.
I design retrieval-augmented generation systems using vector databases, semantic search, and embeddings. I also optimize real-time inference and develop A/B testing frameworks for recommendation and ranking algorithms.
At Pinecone, I developed ranking and recommendation models for large-scale personalization systems. I built scalable pipelines for distributed data processing and model training, and contributed to experimentation frameworks and low-latency production inference.
At Alteryx, I led enterprise data science initiatives involving predictive modeling, forecasting, and customer analytics. I unified data ingestion, feature engineering, training, evaluation, deployment, and monitoring in an automated ML pipeline, and guided junior data scientists and analysts.

