)
Nadav Timor
Researcher,
Weizmann Institute & NYC AI Infra House
Nadav Timor is a research scientist at the Weizmann Institute, focused on scaling and accelerating AI inference. He is pioneering advances in speculative decoding, several of which are now defaults in Hugging Face Transformers, the world’s most popular open-source LLM library—setting new standards for inference efficiency. At the NYC AI Infra House, he leads cross-organization efforts that strengthen the backbone of open-source AI inference. His work bridges hands-on research and real-world deployment, helping make AI systems faster, more efficient, and broadly accessible.
Sessions
-
11-Dec-2025AI & Robotics PlaygroundPanel: The Infrastructure Race - Powering GenAI at Scale