Senior Machine Learning Engineer, Audio Forensics

Forward Digital builds the music distribution and catalog protection platform for independent labels.

authio, our forensic AI music detection product, sets the standard for verifying whether music is human-made or AI-generated. Every track is examined by a 12-model neural ensemble that delivers a forensic verdict (Human, AI or Hybrid), identifies the generating platform and produces evidence-grade results, in under five seconds, with 99.42% accuracy. authio is built for the labels, distributors and streaming platforms that protect the integrity of their catalogs.

We are looking for an engineer who expects to work at the frontier of audio intelligence. This is a fully remote position open to candidates based anywhere in the European Union.

The role

• Advance detection research on next-generation music generators

• Build ensemble fusion, calibration and uncertainty systems for adversarial, ever-shifting data

• Bring state-of-the-art self-supervised audio representations into production

• Engineer GPU-native audio pipelines with industry-leading latency

• Apply a rigorous evaluation standard: sealed holdouts, regression testing, proof before every release

What we expect

• 4+ years of applied deep learning with deep expertise in audio or speech

• Expert-level Python, written to production-grade engineering standards

• Mastery of PyTorch and audio signal processing

• Proven experience in out-of-distribution detection, calibration or conformal prediction

• Uncompromising scientific rigor and exceptional written reasoning

• Fluency with CUDA and/or ROCm environments

Distinguishing experience

• Audio deepfake or anti-spoofing research (ASVspoof or equivalent)

• Work with MERT, WavLM, EnCodec or comparable SSL and neural codec models

• Publications at ICASSP, Interspeech or ISMIR

Why Forward Digital

• Proprietary technology, built and owned end to end in-house

• Dedicated GPU infrastructure engineered for forensic-grade analysis at scale

• Engineering decisions grounded in measurement, never assumption

• Your work shapes how the music industry verifies what is real