Principal Applied Research Engineer
Synthesia · Remote · Europe ·
COMPENSATION
€•• – ••k
Listed on the employer's job board (Ashby)
- Principal level role
- Generative video focus
- London headquarters
- Fully remote possible
- Large-scale training
Join the team behind the world’s leading AI video platform to drive the technical direction of offline video generation. You will own the full stack, from large-scale pre-training to post-training alignment, ensuring research translates quickly into product capabilities.
This senior individual contributor role requires hands-on leadership to resolve complex technical gaps and coordinate across research stages. You will define long-term strategy while personally tackling the hardest engineering challenges in generative video.
What you'll do
- Own end-to-end technical direction for offline video generation
- Bridge pre-training and post-training stages for faster shipping
- Define unified long-term roadmap with research leadership
- Resolve critical technical gaps in architecture and stability
- Accelerate research-to-production handoffs and visibility
- Coach junior researchers toward senior technical execution
- Refine team processes for high-velocity cohesive work
What you'll need
- Proven track record training large-scale video models from scratch
- Experience with multi-node training on millions of hours of data
- Deep knowledge of post-training techniques like RLHF and DPO
- Strong focus on data quality and willingness to question it
- Ability to think strategically while executing hands-on
- Strong cross-functional influence without positional authority
- Leadership style grounded in direct technical involvement
- Obsession with shipping research to users rather than just notebooks
Nice to have
- Owned major model capability jump from training to deployment
- Accountability for both pre-training and post-training outcomes
- Experience with large distributed training runs and compute budgets
- Applied alignment techniques in video or multimodal contexts
- Experience with human feedback pipelines for generative video
Stack
- EXPRESS-Video
- Generative video models
- RLHF
- GRPO
- DPO
How you'll work
- Headquartered in London
- Teams across Europe and the US
- Senior individual contributor position
- Partner directly with research leadership