Sr. Applied Machine Learning Researcher - Generative Models (Remote)

Marketing

Job Description

WHO WE ARE:  

We are a producers playground, delivering music creators the tools they need to bring their ideas to life. With a massive, industry-leading catalog of licensed samples, paired with powerful AI, and access to affordable plugins and DAWs, Splice kicks sound discovery, inspiration, and creative output into overdrive. 

HOW WE WORK:  

At Splice, DISCO is a rallying cry for collaboration, accountability and unity within our organization; Direct, Inclusive, Splice Together, Creator Centric and Optimistic. Our shared success depends on our ability to support one another, work well together and communicate directly. By embracing flexibility and a unified approach, we can navigate anything that’s thrown at us. 

Splice embraces a culture of remote work. You’ll see your colleagues showing up from across the US and the UK. In order to keep us working well as a team, we have regular communication, including Town Halls, departmental All Hands and get-togethers.

When you join Splice, you join a network of colleagues, peers, and collaborators. Are you ready?

THE ROLE:

We are seeking an exceptional Applied Researcher with experience in Generative models for audio using Latent Diffusion, as well as symbolic music generation techniques based on Transformer architectures. Solid experience and track record in only one of the two areas would be considered. The ideal candidate will bring a research-focused mindset with practical application skills, translating state-of-the-art techniques to novel, usable and performant designs and solutions. At Splice, we believe that Generative AI has the potential to augment and extend the sonic boundaries of our human-made, world class catalog, and bring powerful unlocks to our users’ creative workflow.

TEAM INFORMATION:

The Splice AI & Audio Science team is dedicated to pushing the boundaries of artificial intelligence applied to audio data, with the mission to empower music creators everywhere. Being musicians ourselves, we are deeply committed to the use of AI in a creator-centric, ethical and responsible way. Our team consists of passionate and creative individuals who thrive in a collaborative, innovative, and fast-paced environment.

WHAT YOU’LL DO:

  1. Generative AI Research: Conduct literature research and experimentation in the field of ML-based generative audio using Latent Diffusion and symbolic music generation using Transformer-based architectures.
  2. Model development: Collaborate with our ML Engineers to design performant model architectures for efficient ML-based audio synthesis and symbolic music generation, as well as adapting and fine-tuning existing models. Explore, adapt and implement core building blocks for generative models, such as general Variational Autoencoders (VAEs), Neural Audio Codecs (RVQ / VAE), GANs, Diffusion Models, and Transformer-based architectures.
  3. Prototyping: Develop proof-of-concept prototypes to showcase and validate capabilities and use cases using generative audio/symbolic models. Iterate and refine models based on quantitative/qualitative feedback and evaluation metrics.
  4. Collaboration: engage with academic and open source communities to stay up to date with the latest developments in the space, collaborate in joint projects, and identify top talent for our AI & Audio Science team’s future hiring needs.
  5. Stay up-to-date with the latest academic and industrial research in generative models for music, incorporating relevant findings into our applied research and product development processes.
  6. Documentation and Knowledge Sharing: Document research findings, methodologies, and best practices. Collaborate with team members to disseminate knowledge and insights.

JOB REQUIREMENTS:

  • Ph.D. or Master's degree in Electrical Engineering, Computer Science or related Engineering discipline.
  • Background or proven experience in Digital Signal Processing.
  • Proven experience (2+ years) in an Applied Research role focused on Latent Diffusion based generative models for audio and/or symbolic music generation using Transformer-based architectures.. Alternatively, solid experience with diffusion-based models in the image domain, would be considered.
  • Proficiency in Python and deep learning frameworks (e.g., TensorFlow, PyTorch).
  • Familiarity with software development best practices and version control systems (e.g., Git).
  • Strong communication and collaboration skills, with the ability to work cross-functionally with stakeholders in Engineering, Product and Design.

NICE TO HAVES:

  • A relevant portfolio of research projects, publications, or open-source contributions related to generative audio.
  • Prior experience in machine learning model optimization.
  • Background or knowledge in music production.

The national pay range for this role is $165,000 - $206,000. Individual compensation will be commensurate with the candidate's experience.

Splice is an Equal Opportunity Employer 
Splice provides equal employment opportunities to all employees and applicants for employment and prohibits discrimination and harassment of any type without regard to race, color, religion, age, sex, national origin, disability status, genetics, protected veteran status, sexual orientation, gender identity or expression, or any other characteristic protected by federal, state or local laws.