San Francisco, California, United States
Software, data and machine learning engineer with interests toward natural language processing, computer vision and generative machine learning. Also a bedroom music producer.
Building AI agents for the public sector.
Explored creative interests.
Built vector data ingestion pipeline. Shipped end-to-end features utilizing cross-modal retrieval from vector databases. Trained and fine-tuned deep learning computer vision and language models for video search, summarization and classification tasks. Deployed and monitored models in production. Ran manual tests to ensure model quality. Created Weight and Biases workflow to monitor model training. Worked on various MLOps efforts to ensure model quality in inference.
Built and maintained a custom OpenAI Gym environment that simulates Google StreetView using DeepMind StreetLearn images. The environment is then used to train baselines reinforcement learning agents.
Graduate Thesis: Machine Learning for Handling Missing Data in Wearable Electromyographic Systems (Distinction, Highest Honors) Used ML methods to solve missing-data problems in novel fabric-embedded EMG sensors created by the Robot Learning Lab at King's College London.