Audio Data Engineer & Machine Learning Engineer (AI)

TL;DR

Audio Data Engineer & Machine Learning Engineer (AI/Audio): Designing and optimizing deep learning models for speech enhancement and source separation with an accent on digital signal processing, audio data architecture, and evaluation rigor. Focus on building synthetic simulation pipelines and ensuring model reliability in real-world VoIP and recording environments.

Company

hirify.global is building the reliability layer for Voice AI, creating infrastructure to make audio input reliable and measurable for enterprise voice assistants and telephony apps.

What you will do

• Design, train, and optimize deep learning models for speech enhancement, source separation, and sequence-to-sequence audio problems.
• Own the lifecycle of audio datasets, from sourcing and recording to building complex synthetic simulation pipelines.
• Develop sophisticated diagnostics and benchmarking tools to ensure model performance in messy, real-world VoIP environments.
• Improve internal MLOps infrastructure and cloud-based training setups to accelerate iteration.
• Collaborate with Sales and DevRel to translate technical breakthroughs into customer-facing benchmarks and demos.

Requirements

• Deep understanding of digital signal processing (DSP), room acoustics, reverberation, and noise artifacts.
• Expert-level proficiency in Python and PyTorch with a focus on clean, production-ready code.
• 3+ years of experience solving non-trivial ML or Data problems end-to-end.
• Strong scientific mindset with a preference for trustworthy metrics over vanity numbers.
• Startup DNA: comfort with ambiguity and high levels of ownership.

Culture & Benefits

• Competitive salary package with additional benefits and stock options.
• High-impact role in a rapidly growing Voice AI startup backed by top-tier investors.
• Dynamic, fast-paced environment emphasizing ownership and autonomy.
• Opportunity to work with a world-class team of engineers and define the landscape of Voice AI technology.