Conversational Modelling Research Engineer
Tavus · Remote · Remote ·
COMPENSATION
€•• – ••k
Listed on the employer's job board (Ashby)
- PhD or equivalent research exp
- PyTorch + Deep Learning
- Multimodal Model Research
- Hybrid SF/London/Remote
Join Tavus, a Series B research lab backed by Sequoia Capital and Y Combinator, to pioneer human computing through AI Humans. You will work on real-time simulation models that enable machines to see, hear, and respond with emotional intelligence.
As part of the core AI team, you will push the boundaries of foundation multimodal conversational models. Your work will bridge the gap between people and machines, creating trusted agents capable of meaningful, face-to-face interactions at scale.
What you'll do
- Research Large Multimodal Models for conversational avatars and talking heads
- Develop methods to model verbal and non-verbal conversation aspects in real time
- Experiment with fine-tuning and conditioning techniques for audiovisual models
- Partner with Applied ML team to transition research prototypes to production
- Stay current with cutting-edge advancements to define future research directions
What you'll need
- PhD in a relevant field or near completion, or equivalent research experience
- Hands-on experience with Large Multimodal Models and generative language models
- Background in tasks like VQA, audio/video understanding, or speech systems
- Experience fine-tuning VLMs for control, conditioning, or downstream tasks
- Solid foundation in deep learning and strong PyTorch pipeline building skills
Nice to have
- Knowledge of large-scale model training and optimization techniques
- Experience working with duplex-conversational models
- Broader understanding of generative AI across various modalities
- Exposure to software development best practices
- Publications at top conferences like NeurIPS, ICLR, CVPR, or EMNLP
Stack
- PyTorch
- Large Multimodal Models
- Generative Language Models
- AudioVisual Multimodal Models
How you'll work
- Preferred hybrid locations: San Francisco or London
- Remote within US or Europe available for exceptional candidates
- Fast-moving startup environment with experimental culture