Sai Ashish Somayajula

Senior Applied Scientist at Oracle AI

I am a Senior Applied Scientist at Oracle AI, building enterprise data agents that turn structured data into trusted business insight. At the core is an ontology-driven semantic store that maps entities, metrics, relationships, and domain rules onto enterprise data, giving agents the context to understand business intent. This semantic layer is the foundation of our NL2SQL system — the system I co-led to the #1 Archer NL2SQL 2025 win and #1 Spider 2.0 Lite result, two of the most competitive enterprise NL2SQL benchmarks. Building on this retrieval foundation, I focus on agents that reason over evidence, detect patterns, and surface insights proactively, moving toward a future where LLMs help humans understand their data, not just query it.

I hold a PhD from the ECE department at University of California, San Diego, where I was advised by Professor Pengtao Xie and received the 2025-2026 Dr. William S.C. Chang Best Dissertation Award. My thesis investigated how to adapt large language models to specialized, high-stakes domains where data is scarce, distributions shift over time, and output attribution is critical. Working across healthcare and biology, I developed techniques spanning synthetic data generation, continual learning, and watermarking, with the long-term goal of models that domain experts can rely on to reason over complex domain specific data.

I obtained my Bachelor's degree in Electrical Engineering with a minor in Computer Science and Engineering from the Indian Institute of Technology, Hyderabad. I was fortunate to be advised by Professor Sumohana S. Channappayya and Professor Adity Siripuram. I secured the second-highest CGPA in the B.Tech program across all departments and was twice awarded the academic excellence award.

Email CV Google Scholar LinkedIn GitHub

Feel free to reach out to discuss ideas. I'd be glad to connect and brainstorm together.