Staff Research Engineer - Multimodal Generative AI

Jobicy; United States; On-site; USD 200,000 - 260,000; Full time

Job summary

Synthesia in the United States is seeking a Staff Research Engineer to shape the voice team's long-term vision. You will define and own the design and implementation of core components, working with the voice lead and collaborating with video teams and senior staff.

You will develop voice-to-voice models that generate text and voice together, enable back-channeling, and support natural conversations. You will roadmap, implement, and ship outcomes to product.

Qualifications

• Strong understanding of generative modeling for multimodal data.
• Experience with large language models or transformer-based architectures.
• Proven ability to prototype quickly and iterate.

Responsibilities

• Shape roadmap for new model capabilities across audio, text, and video.
• Propose multi-modal system architectures (text and voice).
• Develop streaming and conversational systems with low latency.
• Design solutions for emotional expressiveness and natural interaction.
• Implement designs from pretraining to post-training.
• Ship production-ready models with optimized runtime.

Skills

Generative modelling

Transformer models

PyTorch

Time-series modelling

Prototyping

End-to-end training

Software engineering

Tools

Distributed training

Model optimization

Job description

Synthesia in the United States is seeking a Staff Research Engineer to shape the voice team's long-term vision. You will define and own the design and implementation of core components, working with the voice lead and collaborating with video teams and senior staff.

You will develop voice-to-voice models that generate text and voice together, enable back-channeling, and support natural conversations. You will roadmap, implement, and ship outcomes to product.