Founding AI Research Scientist

The Company
Our client, an applied AI research company building generative audio models, is looking for a talented AI Researcher. Founded by a team from Spotify, Google and Oxford, their mission is to develop AI that can see, interpret and explain our world to ultimately interact with it through sound. The company is unlike any other in this space, with a focus on bolstering the creativity of anyone who works with audio-visual media formats, rather than on replacing human artists. It is a well funded and well backed organisation.
The Role
As a Founding Research Scientist, you will breathe life into novel technology that bolsters the creative processes of anyone working with audio.
You will take the lead on solving real-world challenges inherent to music synthesis, generative modelling, and multimodality while having access to enormous amounts of data and compute.
You will get the chance to build novel foundation models for audio from the ground up, encountering and solving hard problems across research, engineering and product along the way.
Every advancement you make will help unlock value for creators, including some of the largest companies and creator communities in the world
Experience Required:
Must Haves
PhD in AI, Computer Science, or related area with considerable experience in at least one of the following areas: Audio Synthesis, Music Information Retrieval, Generative Deep Learning.
Strong coding skills in Python and a track record of implementing research papers in PyTorch.
Experience with preparing, preprocessing and training on extremely large audio datasets.
Solid first principles understanding of the state-of-the-art in generative audio and their code implementations (Diffusion Transformers, Latent Diffusion, Autoregressive token prediction, VAEs etc.)
Ability to connect research to product milestones, work at high velocity, and commit to effective communication and documentation.
Nice to have
Publications in generative audio in top audio AI conferences such as ICASSP, ICMPC, ISMIR, NIME, ICNMC etc.
Knowledge of music information retrieval and/or audio signal processing (music captioning, audio classification, etc.)
Multi-node distributed training experience (e.g. familiarity with DeepSpeed, Ray, PyTorch Distributed, and/or cluster management)
Compensation and Benefits
Competitive starting salary.
Equity grant to reflect the incredible value you will bring to the company.
Latest and greatest hardware.
Full health, dental, and vision coverage.
Unlimited paid time off. Life in a startup can be intense, and the client appreciates the importance of a healthy work-life balance.
*Please note, you must have the right to work in the UK to be eligible for the position. There is no visa sponsorship offered as part of this role.