Ajinkya Kiran Mulay

I’m a Senior Research Scientist at Meta’s Integrity and Safety organization, where I design and deploy large-scale LLM-driven systems for spam, scam, and adversarial content detection across Facebook, Instagram, and Threads.

Before that I did a Ph.D. in Electrical and Computer Engineering at Purdue (2024), where my research lived at the intersection of sparse recovery, differential privacy, and federated learning. The main output was SPriFed-OMP — a federated algorithm that gets near-SOTA performance at roughly 8–10% of dense parameter counts, under formal DP constraints.

This site is called the pursuits because pursuit algorithms — OMP, FoBa, CoSaMP — are the thread that runs through all my work. Also because “pursuit” is just a good word.