Our lab ran an AI agent hackathon to solve real genomics workflow pain points using ReAct, Plan-and-Execute, Router patterns, tool calling, RAG, and safety guardrails. Explore demos, winners, and project highlights below.

Why an AI Agent Hackathon in a hybrid computational and genomic lab?

Modern genomics research increasingly depends on complex, multi-step computational workflows. This hackathon was designed to help our team internalize agent concepts, planning, tool use, context engineering, and safety, and apply them to real bottlenecks in practical analysis and biological and clinical interpretation.

Explore prototypes built by lab members, including our wet-lab personnel! Each project includes the problem, agent architecture, techniques, safety guardrails, and product demonstration.

🏆 Implementation Excellence Award — Most complete, robust, and usable
Tian Le

🏆 Innovation Award — Most creative agent design or novel technique integration
Gloria (Wenting) Song

🏆 Impact Award — Highest immediate value to daily research workflows
Xusheng Ai

We’re building practical, research-grade automation that respects scientific rigor: transparency, reproducibility, and safety. If you’re interested in agentic systems for genomics, workflow acceleration, or interpretable AI tooling, we’d love to connect.