Principal Applied Research Engineer

At Synthesia we really care about video generation, especially about human centric avatar video generation. This led us to release models such as EXPRESS-Video, and soon our latest video model - these are the best avatar video models in the world, and we are committed to continuing and double down our efforts in leading that area. Our goal is to get to human centric video models that can generate arbitrary long videos at high resolution with arbitrary actions and events. That means continuously training large generative video models from scratch with the proprietary data pipelines and compute infrastructure to support it at scale.