Fields: Machine Learning, Music Information Retrieval, Signal Processing, Computational Musicology
At Hyph, we’re building the future of interactive music creation, a platform that lets anyone generate, remix, and reimagine music intelligently. We’re now opening a few spots for Master’s Thesis students who want to explore the frontier where AI meets music.
We’re flexible in shaping projects around your background and interests, but here are some areas we’re especially excited about:
Music Understanding: automatic detection of key, mode, tempo, sections, and structure from audio stems
Representation Learning: learning stem-level embeddings using VAEs or contrastive learning
Style Adaptation: adapting musical material between different keys and scales while retaining musical identity
Signal Processing Meets ML: hybrid approaches combining DSP feature extraction with deep neural models
Supervision from Hyph’s engineering team
Access to one of the world’s largest private dataset of professionally tagged stems and templates
The opportunity to publish, present, and deploy your work in a real creative product
A fun, fast-paced environment where music meets AI
Background in machine learning, computer science, or audio engineering
Interest or experience in music theory, DSP, or MIR
Solid Python programming and familiarity with ML frameworks (PyTorch/TensorFlow)
Curiosity, creativity, and the drive to build something new
Based in Stockholm, but we welcome students from anywhere in Europe. For exceptional students we can make remote arrangements.
If you are applying as a group, please submit a single joint application.
Our tech stack
Backend: Python, FastAPI
ML: Python, PyTorch
Data: Postgres
Cloud: AWS
Mobile: Flutter
Frontend: Angular (TypeScript)