Turning recorded sound into readable sheet music with AI, built for musicians who think in audio.
Founding Music AI Engineer
You'll design the neural networks that listen to a piano phrase, a hummed melody, or a dense multitrack recording and produce notation, MIDI, and piano-roll visualizations that musicians can immediately use. The core challenge spans polyphonic transcription, source separation, onset detection, and pitch estimation—woven into a product that feels instant. You'll own model architecture, training pipelines, and inference optimization from day one, working directly alongside a small team that ships to real users every week. This is a ground-floor engineering role where your models define the musical fluency of the entire platform.
What they're looking for
- Deep fluency in Python and PyTorch, with experience training and debugging neural networks on real-world audio or sequential data
- Background in music information retrieval, digital signal processing, or related ML domains (automatic transcription, source separation, beat tracking, etc.)
- Comfort owning ambiguous research-to-production problems: you can read a paper on Friday and have a working prototype evaluated by Monday
- Strong software engineering instincts—you write maintainable code, care about latency and memory, and know when to refactor
- 0–4 years of experience (post-degree or self-taught); what matters is demonstrable work, whether that's publications, open-source projects, or shipped products