Director of Data Science, Commerce — Site Lead

About the Role

InMobi's Demand-Side Platform (DSP) operates real-time bidding at millions of queries per second with latency under 50 milliseconds. Currently, the platform is being restructured to focus on commerce outcomes, shifting from impression and click optimization towards driving measurable downstream results such as purchases and return on ad spend (ROAS) for advertisers.

The vision aligns with leading performance-commerce platforms that utilize a unified AI engine to optimize bids directly for conversions and ROAS, transforming the DSP into a customer acquisition channel for commerce advertisers. Achieving this requires innovative modeling approaches, which is the area targeted for significant development.

Key Responsibilities

• Establish and lead a new Applied Science research team based in Beijing to advance frontiers in commerce-DSP intelligence.
• Assume the site lead role providing technical and people leadership, including team building from scratch with full support in budget, headcount, and strategic resources to attract top-tier talent.
• Define and drive the research agenda covering commerce advertising challenges such as bid optimization, auction theory, budget pacing, conversion and ROAS prediction with large-scale models under delayed and sparse feedback, contextual bandits and sequential decision-making for live bidding exploration/exploitation, embeddings and recommendation systems for commerce intent, and incrementality and causal inference for true lift measurement and experiment design.
• Oversee the full lifecycle from research concepts to deploying models in live bidding environments with assistance from Bangalore platform and serving teams, ensuring scalability at millions QPS and sub-50ms latency.
• Hire, develop, and lead a high-caliber applied-science team, setting standards for research rigor, engineering quality, operational cadence, and team culture while providing daily technical and managerial guidance.
• Collaborate closely as part of a unified organization with the Bangalore site, maintaining joint roadmaps, shared standards, clear ownership definitions, and effective communication with product and engineering leadership across locations.

Qualifications and Experience

• Over 15 years of professional experience or equivalent in applied science or machine learning, including extensive experience leading technical teams.
• Proven success in deploying machine learning systems in production at scale, particularly within latency-sensitive live environments with measurable business impact.
• Advanced expertise in one or more domains: auction and bidding mechanisms, large-scale prediction models (CTR, CVR, ROAS), recommendation models and embeddings, and causal inference or incrementality.
• Proficient with modern deep neural network architectures relevant to this field, including deep CTR/CVR networks, sequential and transformer models for user behavior, two-tower retrieval, and multi-task learning for ranking and prediction.
• Experience with reinforcement learning and contextual bandits applied to sequential decision-making in auctions, such as bid optimization, budget pacing, and exploration/exploitation strategies.
• Background in adtech fields such as DSP, RTB, programmatic advertising, and expertise in commerce, retail media, or performance marketing focusing on optimization for purchase and ROAS.
• Strong technical leadership with hands-on abilities to guide direction, assess research, and resolve complex modeling challenges.
• Demonstrated ability to recruit, mentor, and retain senior applied scientists and engineers.
• Excellent communication and collaboration skills for cross-site coordination; fluency in English is required.
• PhD preferred in fields like Computer Science, Machine Learning, Mathematics, Statistics, or Operations Research.