Summary
Shyama Dorbala is a machine learning founder and engineer with a decade of experience building production ML systems across ad tech, analytics, and speech. She led the migration of Liftoff Mobile’s conversion models to a modern deep-learning stack—designing end-to-end training pipelines and deploying models with Spark, PyTorch, and Triton while mentoring cross-functional teams. Prior roles at Google involved predictive modeling for GA4 and building voices for Google Assistant, reflecting a blend of applied ML and speech expertise from her USC graduate work. Her career began in hardware and embedded signal processing (NVIDIA, Oracle, sensor startups), giving her a systems-minded approach to ML infrastructure. Now co-founding Ikeloa in Mountain View, she focuses on conversion modeling and scalable ML products that bridge research and production. Colleagues describe her as technically precise, pragmatic, and curious—aptly summed up by her playful GitHub persona: “a madhatter, with chai-biscuit.”
10 years of coding experience
10 years of employment as a software developer
Masters of Science, DSP, Machine Learning, Speech, 3.88/4.0, Masters of Science, DSP, Machine Learning, Speech, 3.88/4.0 at University of Southern California
Bachelors(Hons in DSP), Electrical Engineering, 3.8 (major), Bachelors(Hons in DSP), Electrical Engineering, 3.8 (major) at Indian Institute of Technology, Hyderabad
French