Senior Applied Research Engineer - Video - remote job

Synthesia is hiring a Senior Applied Research Engineer - Video - a senior role, so the employer expects you to make the calls yourself. The role sits in IT and programming and is fully remote. The company hires candidates based in Europe.

The advert singles out Node.js, so experience with that one tool is what will decide it.

The employer did not publish a figure; that is settled at interview. The advert sets no condition on working hours.

This vacancy passed an automated check: the list excludes any advert requiring a foreign work permit, visa sponsorship, a particular citizenship, or residence in a specific country.

Synthesia stands out as a global leader in AI video technology, serving over 90% of the Fortune 100. Established in 2017, the company is based in London, with operations and teams spread across Europe and the United States.

As artificial intelligence transforms our daily lives and workplaces, Synthesia innovates to boost visual communication and enhance corporate skill development, ensuring teams work more effectively and remain central to thriving organizations.

In our latest Series E funding round, we secured $200 million, bringing our valuation to $4 billion. Our cumulative investment totals over $530 million from top-tier investors like Accel, NVentures (Nvidia’s venture capital arm), Kleiner Perkins, GV, and Evantic Capital, along with founders and leaders from Stripe, Datadog, Miro, and Webflow.

About the Position Join Synthesia as an Applied Research Engineer in our Video Team, where you'll contribute to the development of cutting-edge, production-ready foundation models for human-centric video creation.

You'll be part of a dedicated team that combines expertise in large-scale generative models, distributed systems, and production engineering. Our goal is to create and refine video base models that deliver realistic, controllable, and emotionally expressive synthetic humans on a massive scale. This role is practical and application-focused, not purely academic.

Your responsibilities will include:

- Developing and scaling latent video diffusion models specifically for human-centric video generation. - Enhancing control mechanisms (pose, emotion, script, camera) without compromising quality. - Advancing distributed training strategies (DDP, FSDP, DeepSpeed, sequence parallelism) under practical computational constraints. - Improving training stability across multiple nodes. - Designing robust evaluation frameworks that integrate both automated and human-based assessments. - Optimizing inference processes for minimal latency, high resolution, and cost-effectiveness. - Conducting controlled experiments and interpreting data to guide high-impact modeling decisions. - Upholding rigorous engineering practices: reproducibility, experiment tracking, CI/CD, monitoring.

You'll need to move swiftly, test multiple hypotheses simultaneously, identify key signals early, and prioritize tangible outcomes over extended exploration.

Requirements Essential Skills:

- Extensive experience training deep learning models at scale. - Proficiency in Python and PyTorch. - Hands-on experience with diffusion models, particularly in the video domain. - Background in large-scale multi-GPU/multi-node training. - Knowledge of distributed training techniques (DDP, FSDP, DeepSpeed). - Ability to design and interpret controlled experiments.

Desirable Experience:

- Prior experience with video diffusion models. - History in avatar or human-centric generation. - Familiarity with world/interactive models. - Experience optimizing inference systems for real-world use.

Our Tech Stack

- Programming languages: Python, PyTorch - Frameworks: CUDA, DeepSpeed, distributed training & inference - Tools: Sequence parallelism, AWS, SLURM, Docker - Platforms: GitHub, CI/CD pipelines

What We Look For

- A blend of research-driven curiosity and practical focus. - A strong commitment to delivering results rather than just publishing findings. - Quick exploration of multiple ideas, followed by early elimination of low-priority directions. - Clear communication and scientifically grounded presentation of results. - Independence in execution with effective teamwork.

Why Join Synthesia

- Be part of a fast-growing Generative AI company shaping the future of video technology. - Work on human-centric video projects with real-world impact. - Solve challenging issues in scalability, stability, and controllability. - Influence the evolution of next-generation synthetic human technology. - Thrive in a highly technical, high-ownership environment where your work is implemented.

If you're eager to work on innovative generative video models and see your research applied in practical settings, we'd love to connect.

Our Culture

At Synthesia, we are dedicated to building, not just talking or planning. We seek the brightest, kindest, and most dedicated individuals to focus on their best work with minimal distractions. Our principles guide our decision-making, feedback processes, and work structure to enable everyone to work at their peak. Discover more about our values here.

Our Reach

We serve over 50,000 clients, including 50% of the Fortune 500. Our satisfied customers include Heineken, Zoom, Xerox, McDonald's, and more. Read their stories and reviews on G2.

Proprietary AI Technology

Since 2017, Synthesia has been at the forefront of Generative AI advancements. Our AI technology is developed in-house by a team of top AI researchers and engineers. Learn more about our AI Research Lab and the dedicated professionals behind it.

AI Safety, Ethics, and Security

AI safety, ethics, and security are integral to our mission. While the full impact of AI on society is still being understood, our commitment is clear: People come first. Discover more about our AI Ethics, Safety, and Security initiatives.