Machine Learning Team Lead

Company: Speechmatics
Apply for the Machine Learning Team Lead
Location: Cambridge
Job Description:

Overview

Lead and grow a world-class ML team focused on translating cutting-edge speech research into scalable, customer-centric solutions. You’ll provide technical guidance, ensure robust end-to-end ML development, and align work with business goals to production. The role is central to advancing Speechmatics’ mission to Understand Every Voice by delivering high-quality, deployable models. You’ll work with cross-functional teams in a fast-paced, innovative environment that values learning and impact.

Pay / Benefits

  • Hybrid work with 2-3 office days
  • Focus Fridays
  • Private medical and dental
  • Pension/401K matching
  • Home office equipment allowance
  • Generous holiday allowance

Responsibilities

  • Lead, mentor and develop the ML engineering team
  • Provide technical oversight for end-to-end ML development
  • Guide adoption of software engineering best practices
  • Plan work, allocate resources and coordinate with cross-functional teams
  • Drive technical output to deliver scalable, high-quality ML solutions that meet strategic goals

Key requirements

  • Guiding and mentoring ML engineers
  • Ability to prioritise work by impact, feasibility, and time-to-value with clear success metrics
  • Strong understanding of the modern ML stack (transformers, GQA, KV-caching)
  • Expertise in distributed training techniques
  • Familiarity with inference optimization (dynamic batching, flash attention, speculative decoding)
  • Proven track record of deploying models at scale
  • Proficiency in Python and ML frameworks (TensorFlow, PyTorch)
  • Experience with cloud platforms
  • Familiarity with software engineering practices and ML testing frameworks
  • People leadership
  • Cross-functional collaboration
  • Impact-focused prioritization
  • Transformers and related concepts (GQA, KV-caching)
  • Distributed training
  • Inference optimization techniques (dynamic batching, flash attention, speculative decoding)

…

Posted: October 6th, 2026