Principal Research Engineer
Core
Technical leadership for offline video generation, bridging pre-training and post-training to ship human-centric avatar video models at scale.
Role type
Principal Research Engineer (L7), senior IC with strategy scope
Builds
Foundation models for human-centric avatar video generation
Domain
Generative AI, video generation, multimodal models
Deliverable
production ML models
Required skills
training large-scale generative models from scratch, post-training techniques (RLHF, GRPO, DPO), data quality assessment, strategic multi-year research planning, cross-functional influence, hands-on technical execution
Preferred skills
owning end-to-end model capability jumps, operating large distributed training runs, applying alignment techniques in video/multimodal contexts, leading research teams while remaining hands-on
Technologies
proprietary data pipelines, compute infrastructure, RLHF, GRPO, DPO
Responsibilities
Define end-to-end technical direction for offline video generation, partner with leadership to define long-term roadmap, identify and resolve critical technical gaps in the pipeline, accelerate research-to-production handoffs, coach junior researchers, shape team structure and processes
Seniority
Principal, strategy & mentorship
