Skip to content

Home

Profile

Senior Software Engineer at Google working on Web-Search Ranking and Large Language Models. I hold an M.Tech degree from IIT Bombay and a B.Tech degree from Anna University.

I have previously worked at Yahoo! Research and Alcatel Lucent Labs, building expertise in information retrieval, machine learning, and large-scale distributed systems.

My focus is on solving complex problems at scale—from query understanding and ranking algorithms to training and deploying LLMs that impact hundreds of millions of users daily.


Selected Work

Google Inc, Mountain View, CA

  • Built and maintained query recommendations and ranking systems serving 800+ million daily active users. Developed Recommendations and Retrieval Models that power app discovery and ranking. Designed and deployed an auto-complete suggestion model handling 80K queries per second. Optimized search relevance through machine learning pipelines and A/B testing frameworks.

  • Led development of an on-device universal search system that enables instant app and content discovery directly from the device home screen. Implemented deep neural networks for app predictions, improving user engagement and reducing cold-start latency. Focused on privacy-preserving machine learning techniques for on-device inference.

  • Contributed to core search ranking and understanding systems serving billions of queries daily. Worked on query understanding using NLP techniques, core ranking algorithms, and ScaM (Scalable Machine Learning) Deep-Neural Retrieval systems across ~300M document index. Developed query classification models for better intent understanding and built internationalization systems to support search across multiple languages and regions.

Yahoo Research, Bangalore, India

  • Focused on content understanding and information retrieval for Yahoo Image Search. Developed algorithms for image classification, semantic understanding, and relevance ranking.

Publications

  • Automated LaTeX from Handwritten Math — Developed neural networks to automatically convert handwritten mathematical equations into LaTeX format. Published on arXiv:2412.03853.

  • Protein Crystallization Prediction — Applied machine learning techniques to predict protein crystallization conditions, improving experimental efficiency in structural biology. Published at CIBB.

  • [2015] Opinion Summarization using Submodular Functions — Published at EMNLP with a Best Reproducibility Award for open-sourcing code and detailed experimental setup.

Misc

  • Active conference reviewer for top-tier venues: ACL, EMNLP, IEEE TNNLS
  • 20+ Google Awards for innovation, technical excellence, and impact on products serving billions of users