LangChain
Observability and evals. Learn more →
Jun 2025 - Present
Columbia University
Investigating why models fail on complex programming tasks under Prof. Baishakhi Ray. Built classifiers to distinguish between reasoning and output errors in AI systems.
Jan 2025 - Jun 2025
Columbia University CS
Led workshops on Natural Language Processing, Computer Vision, and Machine Learning for 100+ students in an applied AI course.
Columbia AI Alignment Club
Conceived and developed a novel mathematical framework and benchmark dataset for detecting cumulative risk in AI agents, securing Open Philanthropy funding.
Pika
Fine-tuned image-to-video models at a generative video company. Built software to automate video pre-processing workflows and researched alternative training pipelines.
Jul 2024 - Jan 2025
