Vincent Tu is an M.S. Computer Science researcher and engineer focused on LLMs, language agents, and computer-use reasoning, currently interning at Collinear AI and formerly driving agent research at Simular. He leads the open-source Agential project, building a plug-and-play Python toolkit and CI-backed experiments that standardize agent benchmarking and accelerate reproducible research. His work at Simular delivered substantial benchmark gains (e.g., +32.7% on OSWorld) and produced a compositional agent framework under review at COLM 2025. Vincent combines academic research at UCSD’s Q-Lab with hands-on MLOps experience—deploying LLM-based systems using Hugging Face, W&B, Docker, Kubernetes, and AWS—and has shipped production ML pipelines for tabular credit models. Prior contributions span biological computer vision tools (SLEAP/dreem) and technical writing for Weights & Biases, reflecting a habit of translating research into practical, community-facing artifacts. He’s notable for bridging rigorous benchmark-driven research with pragmatic tooling that helps teams reproduce and deploy agentic systems.
6 years of coding experience
2 years of employment as a software developer
University of California, San Diego
GPA (UW/W): 3.98/4.65, Rank: 6/409, GPA (UW/W): 3.98/4.65, Rank: 6/409 at Gabrielino High School
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.