Principal Machine Learning Engineer

We are hiring a Principal Machine Learning Engineer to work on cutting-edge R&D and translating research into customer-centric solutions.

As innovators in speech technology, our mission is to Understand Every Voice—a vision that has propelled us to be world leaders of Voice AI. Fuelled by innovation, inclusivity, and a passion for making a global impact through world-leading Speech AI, we're looking for an experienced Principal Machine Learning Engineer to accelerate our efforts towards exceptional speech solutions.

Our Modelling Team trains diverse models, including large self-supervised ones, supporting Speechmatics towards being the most accurate speech recognition system globally. It also ensures their deployment into production, working with the latest developments in ML, but also with the best engineering practices for software engineering and model serving.

This is a hands-on, technical leadership role with high ownership. You will develop and deploy advanced speech systems that power our products as well as co-define the technical vision for ML, drive innovation, and mentor engineering teams. Your work will span the entire codebase.  

What You’ll Do:

• Develop and deploy ML models, translating research into scalable, maintainable code and services
• Optimise ML models for speed, accuracy, and cost efficiency
• Evaluate and integrate cutting-edge approaches into our ML stack
• Identify and solve complex ML problems across the organisation
• Define and enforce best practices for code quality, model lifecycle management, etc.
• Mentor engineers and foster a culture of technical excellence and innovation
• Co-define the long-term technical vision

What We’re Looking For:

• Deep understanding of the modern Machine Learning stack, for example:
  • Knowledge of contemporary transformer architectures (e.g., GQA, KV-caching) and best practices
  • Expertise in distributed training techniques
  • Familiarity with optimisation strategies for model inference (e.g., dynamic batching, flash attention, speculative decoding)
• Proven track record of developing and deploying models at scale
• Expertise in Python, ML frameworks (TensorFlow, PyTorch), and experience with cloud platforms
• Strong familiarity with MLOps principles, CI/CD pipelines, and containerization (Docker, Kubernetes)

  With preferred backgrounds covering some of the following:

• Experience with speech-to-text or text-to-speech
• Contributions to open-source projects or published research in top-tier conferences