SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Synthesia is a $4B AI video platform used by over 90% of the Fortune 100. The company is seeking a Staff Research Scientist to join a 40+ person R&D team focused on generative AI and avatar-centric interactive video diffusion models.
In this role, you will set research direction and roadmap for dyadic interaction modeling—balancing long-term research bets with near-term product impact. You'll advance the state of the art in the perceptual layer of interactive agents, including understanding user audio and video and generating contextually appropriate reactions. Key responsibilities include:
- Setting research direction and roadmap for dyadic interaction modeling
- Advancing the perceptual layer of interactive agents (audio/video understanding and reaction generation)
- Post-training multimodal models to generate natural dyadic interactions from user audio and video inputs
- Adapting diffusion models to new conditioning signals such as conversational state, turn-taking, and listener cues
- Building robust evaluation frameworks and test suites for interaction quality
- Partnering with the data team to define data needs and shape high-quality datasets
- Leading and mentoring a small group of researchers and driving technical decisions across research, data, and engineering
The team moves fast, iterates often, and turns breakthrough research into shipped product capabilities. You'll work on models that enable AI agents to hold natural, real-time conversations—understanding what users say and how they say it, then responding with appropriate turn-taking, backchannels, gaze, and facial and body reactions.
Synthesia was founded in 2017 and is headquartered in London with offices across Europe and the US. The company recently raised $200M in Series E funding at a $4B valuation from premier investors including Accel, NVentures (Nvidia's VC arm), Kleiner Perkins, GV, and Evantic Capital.
REQUIREMENTS:
- Deep ML expertise and extensive hands-on experience with diffusion models, ideally for video or avatar generation
- Strong publication record at top-tier venues (CVPR, ICCV, ECCV, NeurIPS, ICML, ICLR, SIGGRAPH) in areas such as world models, dyadic interaction, or video diffusion, or equivalent demonstrated impact
- Experience leading a small team of researchers and mentoring junior members
- Track record of taking research from idea to production
- Proficiency in PyTorch and modern ML tooling for large-scale training
- Clear communication of hypotheses, experiments, and results; ability to influence direction across teams
NICE TO HAVE:
- Experience with real-time or streaming generation, including autoregressive video diffusion
- Distillation or other techniques for low-latency inference
- Audio-driven facial, gesture, or full-body motion modeling
- Conversational modeling (turn-taking, backchanneling, listener-response generation)
