A distributed-systems engineer who founded and led several 0-to-1 infrastructure efforts at Meta, from a DAG executor running multi-billion workflows a day to Meta's internal agent memory platform.
Lionell Loh spent six years at Meta founding and leading infrastructure efforts from scratch, then scaling them across the company and to the billions of people who use Meta's apps: a video orchestrator that underpins Meta's video infrastructure and peaked at 6 billion workflows a day; Meta-wide multimodal batch inference for Llama research; and Memex, an agent memory platform with more than 100 use cases. He graduated valedictorian of the Singapore University of Technology and Design (SUTD) and has published research on defending machine-learning models against attack. At Noetive he works on the agentic ecosystem, ML and system infrastructure.
