Software Engineer · AI Systems · Multimodal ML
Co-developer of VANTAGE-Bench — an open benchmark for evaluating vision-language models for Physical AI.
35,027 expert annotations · 3,346 multimodal assets · 4 reasoning pillars
Featured in NVIDIA's GTC Taipei Cosmos-3 keynote
VANTAGE-Bench · Leaderboard · GitHub Multimodal benchmark and evaluation infrastructure for frontier vision-language models, spanning distributed inference, automated scoring, dataset pipelines, and a public leaderboard.
PromptPact Open-source infrastructure for prompt regression testing and LLM evaluation.
MuseBoard Multimodal pipeline for turning visual inspiration into structured marketing intelligence.
AI Systems · Multimodal ML · LLM Evaluation · Distributed Inference · AI Infrastructure
