Research Lead

Alan Dao is an AI researcher and builder passionate about creating systems that push the boundary of intelligence and usability. His work spans multimodal models, robotics, and agentic workflows—focusing on real-time reasoning, spatial understanding, and open-source accessibility. He is the creator and co-author of several impactful systems and research works, including: Jan-nano, a 4B tool-using language model designed for search and question-answering agents Ichigo, a realtime voice assistant using early fusion multimodal modeling AlphaMaze and AlphaSpace, frameworks that bring spatial and symbolic reasoning to LLMs for visual tasks and robotic manipulation ........ His research has been published in major forums like arXiv and Interspeech 2025, where his recent work Speechless explores instruction-tuned models for low-resource languages—without requiring any speech. Alan’s open-source contributions have helped shape the evolving agentic ecosystem, with systems like Nitro (now Cortex) powering efficient LLM inference at scale. His projects are available on GitHub, and his thoughts are shared on Twitter, Threads, and his blog alandao.net.