I am a researcher and engineer at Meta Superintelligence Labs (MSL). I specialize in the development of multimodal large language models (LLM) and was a core contributor to Llama 3.1 and Llama 4. Previously, I was at Facebook AI Research (FAIR), working on natural language processing, speech processing and audiovisual understanding. I published 50+ papers at top-tier AI conferences such as NeurIPS, ICML, ACL and INTERSPEECH.