Synthesia is a London-founded AI video company used by businesses to create training, communication and marketing videos with synthetic presenters and voices. In January 2026, the company raised $200 million at a $4 billion valuation.
- Develop expressive, real-time synthetic voice technology
- Work on speech-to-speech systems, neural codecs and generative audio models
- Remote role open to candidates based in Europe
The role
Synthesia is looking for a Senior Research Engineer to improve the voice technology behind its AI video platform. The team is working on voices that sound natural, expressive and responsive enough for real-time use.
This is an applied research position. You will train new models, run experiments and help move successful work into a product used by companies around the world.
The work includes:
- Training and evaluating speech generation models from end to end
- Developing low-latency voice and speech-to-speech systems
- Improving control over emotion, prosody, speaker identity and delivery
- Working with neural codecs, diffusion models and flow-matching methods
- Using quantisation, pruning or distillation to make models faster and more efficient
- Collaborating with research, engineering and product teams to ship improvements
Who Synthesia is looking for
You should have strong experience training deep learning models and a solid understanding of modern generative modelling. The team is interested in people who have worked on speech synthesis, audio generation, neural codecs or closely related areas.
Strong Python and PyTorch skills are important, along with experience using distributed training systems. Evidence of original research is useful, but the ability to build reliable software and work in a shared research codebase matters just as much.
Why apply
This role sits at the point where generative audio research becomes a product. The models need to sound convincing, run quickly and work consistently for real users, which gives the engineering work a clear purpose.
Synthesia lists competitive pay, stock options and a bonus. The company also offers 25 days of annual leave and supports remote work across Europe, with office options in cities including London, Amsterdam, Zurich and Munich.