Summary
Karthik Sankararaman is an AI research scientist with a PhD in sequential decision making and over a decade of experience building RL, bandit, and frontier-model systems that ship at scale. At Meta Superintelligence Labs he leads post-training and RLHF research for Llama/MetaAI, driving technical direction across reward modeling, tool use, agentic behavior and a data flywheel that has powered every Llama release since 2023. His background building RL and contextual-bandit systems for ads, recommendations, and integrity grounds his frontier-model work in real product impact and measurable metrics. Karthik has published foundational papers on constrained bandits and causal robustness, and has a long history of translating algorithmic advances into production systems. Based in Milpitas, he blends deep theoretical expertise with pragmatic engineering to improve model capability, reliability, and alignment in ways that reflect how people actually use AI.
10 years of coding experience
3 years of employment as a software developer
Indian Institute of Technology Madras
Doctor of Philosophy (Ph.D.) Computer Science, Doctor of Philosophy (Ph.D.) Computer Science at University of Maryland
Tamil, Kannada, Hindi, English