Dr. Thomas Prätzlich

I am an Audio AI & Data Science Consultant helping startups, R&D teams, and music-tech/audio-product companies build, evaluate, and improve machine learning systems for audio, music, sound, and speech.

I support clients with technical advisory, prototyping, model evaluation, data strategy, applied R&D, and the translation of research ideas into reliable product features. My consulting focuses on case-specific technical guidance — the kind of specialist input that generic AI tools and broad ML advice often miss.

I hold a PhD in Music Information Retrieval from FAU Erlangen-Nürnberg, with research conducted at the International Audio Laboratories Erlangen, a joint institution of Fraunhofer IIS and FAU. My research covered music synchronization, segmentation, interference reduction, classification, fingerprinting, acoustic similarity analysis, singing voice detection, polyphonic note recognition, and source separation. I have published at leading conferences including ICASSP and ISMIR and regularly serve as a reviewer for ISMIR.

I also bring practical startup experience from music learning, audio ML, and generative music applications, including work on nootone, a smart music sketchbook app developed during The Sound of AI Accelerator.

Available for advisory work, contract research, prototyping, applied R&D, and hands-on audio AI consulting.