ML Data & Platform Engineer — Scale Training Pipelines

Speechmatics; England; Hybrid; GBP 75,000 - 110,000; Full time

Benefits offered by this job

Private Medical

Dental

Global opportunities

Generous holiday allowance

Pension matching

Home office allowance

Parental leave

Job summary

Speechmatics in the United Kingdom seeks an ML Data & Platform Engineer to own the data pipelines and the platform that trains, evaluates, and serves models in production. You’ll work across data infrastructure, distributed systems, and production ML, taking end-to-end ownership to reduce friction and speed deployments.

The role is hybrid with 2-3 office days per week, offering opportunities to impact model performance and production reliability while collaborating with a global ML team.

Qualifications

• Proficient in Python and SQL with backend/data-engineering foundation.
• Experience with Docker, Kubernetes and major cloud provider.
• Experience building data pipelines and ETL/ELT at scale.
• Solid understanding of the ML lifecycle from data to model training, evaluation, and serving.
• Experience with data quality practices and observability in production.
• Ability to design scalable, resilient distributed architectures.
• MLOps experience including model serving, experiment tracking, GPU/training optimization.
• Self-starter mindset; can identify problems and drive fixes without detailed direction.

Responsibilities

• Design, build, and maintain scalable data pipelines for ingesting, transforming, validating, and storing large datasets used to train our models.
• Developing and maintaining web scraping and data acquisition solutions to keep training datasets fresh, high-quality, and available at scale.
• Build and operate the infrastructure that lets the ML team deploy and evaluate new models quickly, and that serves models efficiently and reliably in production.
• Optimising infrastructure for both iteration speed and production reliability, including GPU utilisation, job scheduling, and training efficiency.
• Implementing observability (monitoring, logging, alerting) across data pipelines and ML systems to catch issues early and keep things running smoothly.
• Troubleshooting complex issues across distributed systems, spanning data infrastructure, training, and inference.
• Continuously improving our data and MLOps practices, and helping shape the roadmap for how our platform evolves as we scale.

Skills

Python

SQL

Data pipelines

ETL/ELT

MLOps

Observability

Cloud

Distributed systems

Self-starter

Tools

Docker

Kubernetes

Job description

Speechmatics in the United Kingdom seeks an ML Data & Platform Engineer to own the data pipelines and the platform that trains, evaluates, and serves models in production. You’ll work across data infrastructure, distributed systems, and production ML, taking end-to-end ownership to reduce friction and speed deployments.

The role is hybrid with 2-3 office days per week, offering opportunities to impact model performance and production reliability while collaborating with a global ML team.