← Back to all jobs
Research Scientist - Audio Codec
Mirelo AI
- Employment type
- Full-time
- Country / Region
- Germany
- City / Location
- Berlin, Stuttgart
- Workplace
- Hybrid
- Seniority
- Senior (5-10 years XP)
- Salary
- Not communicated
- Language(s) required
- English
- Job offer language
- English
- Applications open until
- 28 October 2026
Mirelo AI is hiring a Research Scientist - Audio Codec in Berlin, Germany. This is a full-time, hybrid position in Software, Tech & Data Analytics at senior level. Applications are open until 28 October 2026.
More: Music jobs in Berlin · Music jobs in Germany
About the Role
Work at the core of Mirelo AI's modeling stack, designing, training, and evaluating neural audio codecs that directly shape the performance of the company's next-generation music and sound models.
Key Responsibilities
- Develop and implement new neural audio codecs for sound, music and speech optimized for generative model use cases
- Address the specific challenges of using a codec as a latent representation in generative audio models
- Explore trade-offs between continuous (diffusion) and discrete (autoregressive) audio representations
- Develop benchmarking pipelines for codec evaluation
- Conduct experiments with generative models to verify new candidate codecs are useful for downstream tasks
Requirements
- Strong background in deep learning for audio: neural codecs, source separation, speech models, or generative audio systems
- Hands-on experience designing and training neural audio codecs
- Solid understanding of audio signal processing fundamentals
- Strong research and/or open-source track record in audio ML
Details
Hybrid role based in Berlin or Tübingen, Germany. Mirelo AI recently closed a $41M Seed round co-led by Andreessen Horowitz and Index Ventures.
Ready to apply?
Create a free account to unlock the application link for this role.
Sign up / log in to apply