Work

Experience

2026 —

Staff Software Engineer

Research and infrastructure.

2026

Evals Lead

Joined as part of the Groq–NVIDIA licensing agreement; worked on inference and evaluation systems.

2024 – 26

Head of Evals

Led the evals team and the benchmarking behind Groq's model quality work. Created openbench there and open-sourced it.

Projects

2025

An open-source framework for running language-model benchmarks reproducibly, across providers — 30+ evaluation suites behind one interface.

2024

An evaluation framework using debate simulations to assess AI models' reasoning and communication skills.

2024

A multimodal benchmark for testing vision capabilities and reasoning in AI models.