Max Graf
Music AI researcher specialising in deep learning for music analysis, generation, and interactive music systems. Currently ML Engineer for the EU-funded CORPUS project. PhD in AI & Music from C4DM at Queen Mary University of London, with experience building generative models, music metadata pipelines, XR musical instruments, and AI-powered tools for sample exploration. Strong background across PyTorch, audio ML, real-time interaction, and creative music technology.
Core Skills
- Music AI / Audio ML: music generation, MIR, metadata extraction; diffusion, language model (LM), and state-space (SSM) architectures; source separation, semantic search
- ML Engineering: Python, PyTorch, Lightning, scikit-learn, NumPy, SciPy, pandas, TensorFlow, LibTorch
- Creative Tech: Unity, C#, C++, XR/MR, real-time interaction, JUCE, Max/MSP, Pure Data
- Software: backend APIs, full-stack web, JavaScript, Node.js, Svelte, Git
Experience
March 2025 - present
ML Engineer
CORPUS
- Designed, implemented and tested a comprehensive music metadata extraction and analysis pipeline for large-scale music understanding and generation used in the CORPUS Contribution App.
- Developed and evaluated real-time audio-domain music generation models: combined Google/DeepMind’s SpectroStream neural audio codec with a MusicGen-style architecture to train custom models for real-time, high-quality instrumental music generation.
- Trained and evaluated diffusion-based models for controlled, longer-form music generation, using Stable Audio as a baseline and applying inference-time optimisation for steering outputs via pitch, rhythm, and energy.
- Built CORPUS Music Intelligence, an interactive music discovery and generation platform built using a combination of rich music metadata and CLAP-style models for discovery, and audio-domain diffusion models for music generation.
- Full-stack integration of the above ML applications into production applications (backend, APIs and frontend workflows).
September 2024 - January 2025
Researcher & Developer
Netz XR Instrument, Queen Mary University of London
- Secured Innovate UK grants to develop an XR musical instrument combining real-time gesture recognition and expressive sound synthesis (spun out from PhD research).
- Drove prototype iterations, user surveys, and public demonstrations as part of grant–funded commercialisation efforts.
May 2023 - February 2026
Researcher & Developer
WavNav / Audio Maps, Queen Mary University of London
- Developed WavNav, an AI-powered sample explorer using CLAP embeddings, audio feature extraction, semantic search, query-by-example retrieval, and visual sample mapping.
- Built signal processing and feature extraction pipelines for sample analysis, search, and organisation.
- Designed interfaces for fast exploration of large local music collections, focused on music-production workflows.
October 2021 - February 2022
Teaching Assistant
Queen Mary University of London
- Supported the postgraduate module Interactive Digital Multimedia Techniques, delivering workshops on real-time audio interaction, gesture control, and XR prototyping.
2016 - 2019
Software Engineer
RSIT, Vienna
- Part-time full-stack development and data visualisation on medium-to-large projects (JavaScript, Python, Java, C++).
Earlier experience: software engineering internships at ASFiNAG, Frequentis, and Roombonus (2012–2016).
Education
PhD, AI & Music — Queen Mary University of London, 2020–2026 Supervisors: Mathieu Barthet & Andrew McPherson
MSc, Sound and Music Computing (Distinction) — Queen Mary University of London, 2019–2020
BSc, Media Informatics & Visual Computing — TU Wien, 2015–2019
Selected Publications
- Graf & Barthet. Multimodal Hand Tracking for XR Musical Instruments Using Electromyography. LNCS, 2025. Paper
- Graf & Barthet. When XR Meets AI: Integrating Interactive Machine Learning with an XR Musical Instrument. AES International Symposium on AI and the Musician, 2024. Paper
- Graf & Barthet. Combining Vision and EMG-Based Hand Tracking for Extended Reality Musical Instruments. CMMR, 2023. Best Paper Nomination. Paper · Code
- Graf & Barthet. Reducing Sensing Errors in a Mixed Reality Musical Instrument. ACM VRST, 2023. Paper
- Graf & Barthet. Mixed Reality Musical Interface: Exploring Ergonomics and Adaptive Hand Pose Recognition. NIME, 2022. Paper · Code
- Graf, Opara & Barthet. An Audio-Driven System for Real-Time Music Visualisation. AES 150, 2021. Paper · Code
Grants & Awards
- Innovate UK ICURe Explore, £35,000 (2024) — commercial validation of Netz XR instrument.
- Innovate UK ICURe Discover, £3,500 (2024).
- Impact Funding, £10,000 each for Netz (2024) and Audio Maps (2023).
- MIDI Innovation Award for the Netz XR musical instrument (2023).
- Austrian awards for excellent foreign postgraduate/doctoral students (2020, 2021).