I'm a product lead working on AI safety systems for foundation models at Microsoft CoreAI. Outside of work, I’m building cotrust.ai, a community and platform focused on model behavior design, evaluation and engineering.
Previously, I helped ship the first open-source 176B LLM at Hugging Face and led global AI research and engineering communities such as MLT.
Linguistics PhD with research on computational approaches to affect, pragmatic inference and nuance in language models.