- Drive the vision, strategy, and roadmap for personalization in Gemini models.
Engage with consumers, developers, and enterprise customers, and track competitive and market intelligence to identify high-value use cases and capability gaps.
- Establish and scale human and synthetic data, simulated environments, user
personas, and user simulators needed to evaluate and train personalized behaviors, along with programmatic quality control that ensures a high yield of high quality data.
- Own the quality bar for personalization/personalized agent task success by
defining the evaluations, benchmarks, autoraters, and human-rater programs that measure it, and use them to drive model hill-climbing, checkpoint selection, and release decisions. Work with Privacy, Trust & Safety, Security, and Responsible AI teams.
- Partner with post-training teams to set capability targets and close quality
gaps/integration surfaces.