Lead strategy and technical direction for generative and multimodal vision systems. Build end-to-end model workflows, improve realism, controllability, and robustness, drive evaluation and production-readiness, partner with engineering to deploy optimized inference, and mentor scientists while owning a major technical pillar from research to production.

Lead development of map-grounded world foundation models for autonomous driving: train and deploy generative AI (diffusion, transformer, generative video) integrated with simulation platforms, define synthetic data quality frameworks, run PoCs with partners, and translate research into production validation and businessroadmaps to improve perception and planning performance.