Summary
Ajinkya Tejankar is a Founding Research Engineer and PhD candidate specializing in self-supervised and multi-modal (vision+language) representation learning, with 11 years of engineering and research experience bridging academia and industry. He has multiple first-authored papers at top conferences and two Meta AI internships where he probed biases and defended self-supervised models from backdoor attacks. Recently he moved into production-focused ML engineering—optimizing LLM inference (multi-LOra serving, speculative decoding, FP8 quantization) and designing agent evals, simulations, and memory for code-handling systems. Comfortable shipping end-to-end systems from research prototypes to production, he also brings early full-stack product experience from a prior software development career. A not-obvious strength: he leverages simple, pragmatic insights (e.g., bag-of-words for zero-shot and targeted pseudo-labeling) to yield outsized empirical gains. Based in Davis, CA, he combines deep technical rigor with a product-oriented mindset for deploying robust, scalable AI.
11 years of coding experience
10 years of employment as a software developer
Doctor of Philosophy - PhD, Doctor of Philosophy - PhD at University of California, Davis
UMBC
Bachelor of Technology - BTech, Computer Engineering, Bachelor of Technology - BTech, Computer Engineering at Savitribai Phule Pune University