Teams at InMobi

Develop and optimize large-scale machine learning models, open-source LLM infrastructure, and training and inference systems for advertising applications. Build distributed data and model pipelines, adapt DeepSpeed and OpenRLHF, optimize GPU/TPU serving for latency and cost, monitor model performance, and collaborate with applied scientists on research. The role focuses on recommender systems, personalization, targeting, reinforcement learning, and ad intelligence at massive scale.

Conduct research and develop mathematically rigorous algorithms for real-time auctions, dynamic pricing, bid shaping, pacing, traffic allocation, and fraud detection. Apply machine learning, reinforcement learning, online learning, game theory, forecasting, and Bayesian methods. Deploy models at massive scale, run rapid experiments, evaluate long-term system behavior, collaborate with product and engineering teams, and contribute through publications and internal research seminars.

Develop machine learning and personalization systems for real-time advertising at planetary scale. Conduct architectural research on attention-based and post-attention models, evaluate hypotheses offline and online, improve recommender and targeting systems, and transition successful research into production. Build model implementations and data pipelines, guide infrastructure investments, collaborate with engineering and product teams, and contribute to the research community through publications, talks, and conferences.