Hello!¶
I am an experienced AI engineer and consultant with background in industry research labs and applied AI produts. I specialize in bridging the gap between cutting-edge AI research and practical solutions, with experience ranging from big tech to early-stage startups. I'm based in NYC.
Looking for collaboration or need advice? Let’s connect!
Some Projects I've Been Tinkering With Lately¶
ActionWeaver: An open source framework for building LLM applications that simplifies function calling with LLMs. View on GitHub
AutoCoder:A weekend side project: a description-to-pull-request bot that answers questions and makes code changes in GitHub repositories using natural language instructions. Built with ActionWeaver and powered by LLM function calling. GitHub Repository | Watch Demo
AI Agent for immigration lawyer I built a demo for a startup showcasing an end-to-end workflow to automate tasks for immigration lawyers, streamlining the visa application process. Watch Demo
An Empirical Study of Neural Network Training Dynamics: An attempt to uncover what’s really happening during neural network training. Read the blogpost
When does loss-based prioritization fail?: A research paper from my time at Uber, exploring methods to robustly accelerate neural network training. Arxiv
Geospatial Monitoring at Scale: A project while working at Uber - a real-time system that monitors on traffic patterns everywhere Uber operates with Apache Flink and Complex Event Processing. I gave the presentation at Flink Forward San Francisco 2020 on behalf of Uber.
Watch Conference Talk (Flink Forward San Francisco 2020)