LLMs + inference + infra systems + Ray

Data systems for Agents / LLMs (DB group). Advised by Prof. Eugene Wu and Prof. Kostis Kaffes.

Chain-of-thought reasoning, argument extraction / NER and generative summarization for legal corpora. Advised by Dr. Kenneth Marino (DeepMind) + Prof. Kathy McKeown (Columbia).

Building Workers AI - https://ai.cloudflare.com/ - 50ms from 95% of the internet.

(Extended into Fall - while continuing studies)

ML infra, LLM serving, fine-tuning, profiling, benchmarks, evals, inference optimization, scaling backend systems, reading papers, and more.

ML Infra, Search/Data Infra, Privacy, Exascale Tooling and Compute Wrangling

ETL + migrating an Exascale Hadoop cluster to Azure.

https://www.linkedin.com/blog/engineering/open-source/the-exabyte-club-linkedin-s-journey-of-scaling-the-hadoop-distr