Research Engineer - Generative modelling

Are you passionate about pushing the boundaries of AI technology?

Join Anam AI, who are redefining what is possible with real-time, conversational AI personas (aka photorealistic avatars). We are seeking you, an experienced research engineer with expertise in generative models to join our team and help us revolutionise the future of AI personas.

Many current AI avatars are slow, scripted, dubbed, lack the ability to understand or convey emotion and so it's impossible for the interaction to feel natural. Our mission is to change that. Founded by former Synthesia team members, Anam.ai is developing the next interface for technology - real time personas who are capable of understand verbal and non verbal cues and respond with in 500ms. The conversation will feel as natural as interacting with a human.

As a key member of our team, your work will play a crucial role in the development of this groundbreaking technology. You’ll collaborate closely with our CTO and co-founder to tackle ML challenges in generative modelling and seq2seq.

You will take full ownership of the ML R&D pipeline for generative computer vision (including audio-to-video), from problem statement to deployed model. This involves all stages from research, staying current with the latest developments and contributing to internal paper reading group meetings, to data curation & processing, building & training models, and deployments. With our extensive proprietary dataset of hundreds of thousands of talking head videos, you’ll have access to the ideal data to train your models effectively.

Your background should include experience in ML and expertise in implementing seq2seq models. Ideally, you will be skilled in computer vision topics such as lip-sync, talking heads, audio-visual integration, face generation, and face rendering. Familiarity with key architectures such as Transformers, Diffusion models, GANs, and VAEs is essential, along with knowledge of cutting-edge research aimed at real-time and state-of-the-art generative vision.

While research skills are vital, including relevant published papers at top conferences, you should also be a strong engineer, proficient in Python, and experienced in deploying complex models in a commercial setting.

Additional desirable skills include experience with multimodal generative models and distributed computing, specifically running GPUs in the cloud and addressing inference challenges for real-time systems.

Anam AI is a newly founded startup with pre-seed funding of £1.75 million. Joining us at this pivotal moment means your work will significantly impact the future of our product. You’ll be part of a dynamic team of engineers (ex-Synthesia, ex-Twitter, ex-Google) building next-generation multimodal systems, including TTS, ASR, LLMs, Vision Generation, and more. We are also supported by an exceptional industry advisor, an ex-Stability AI executive.

Whilst candidates in London are preferred. Remote work is considered for exceptional candidates, who match all the requirements.

If you are open to relocating to London in the future, visa support can be provided.

Your salary will be dependent on experience and ranges between £90,000 - £120,000. As an early hire, you will also receive generous equity to complement your salary. We offer a 27-day holiday allowance (plus bank holidays) and a standard pension. A bonus structure will be introduced soon.

There are ample opportunities to advance your career here, whether you prefer an IC role or aspire to be a future leader.

Interested in learning more? Please apply via this job advert or contact Marc Powell at Techire AI, talent partner of Anam AI.

Please note Techire AI is a retained staffing partner of Anam AI and all applications must go through Techire AI to be considered.

We do not accept submissions from other recruitment agencies, consultancies, or outsourcing companies.