Sudarshan Srinivasa Ramanujam is a lead engineer at LinkedIn working on Feed ranking and retrieval modeling, where he plays a key role in building and scaling production models that power the LinkedIn Feed. His work centers on applying LLMs to recommendation models and expanding the capabilities of the retriever models that shape the professional content over 1 billion LinkedIn members see.

He is the co-lead author of Large Scale Retrieval for the LinkedIn Feed using Causal Language Models and Connected Content Retriever: Dense Graph Edge Features for Pre-Ranking at LinkedIn. Sudarshan presented the LLM embeddings powering LinkedIn's Feed and video surfaces at CIKM 2025 and serves on the Program Committee for leading conferences including SIGIR, RecSys, and CIKM. He has delivered guest lectures to graduate students at UNC Charlotte on applying LLMs to recommendation systems in industry. He served as head TA for a graduate-level deep learning course at UC San Diego, where he earned his master's in computer science.