Python Software Engineer ML Tooling

Principle HR; Greater London; On-site; GBP 98,000 - 110,000; Full time

Job summary

A leading recruitment agency is looking for a Python Engineer to build and enhance tools crucial for machine learning research. The role necessitates 4+ years of experience in building software, especially with Python, and familiarity with ML frameworks like PyTorch or TensorFlow. You'll work on-site in a hardware-heavy environment, where communication with researchers is key to refining their tooling needs. A strong hands-on engineering mindset is required to address real-world systems challenges effectively.

Qualifications

• 4+ years commercial experience building software in Python.
• Experience with PyTorch or TensorFlow in production environments.
• Familiarity with large/complex ML datasets and workflows.

Responsibilities

• Build and maintain scalable Python tooling for ML research workflows.
• Fix bugs, optimise performance, and implement iterative enhancements.
• Work closely with researchers to gather requirements and translate into software.

Skills

Building software in Python

Hands-on with PyTorch or TensorFlow

Experience with distributed compute

Good communication skills

Tools

Linux/Windows scripting

CI/CD tools

Job description

If you like building Python systems that make machine learning research actually work at scale - this one's for you. You'll develop and improve the internal tooling that researchers rely on daily: distributed compute workflows, data pipelines, and robust infrastructure that connects research code to real environments (including hardware/prototype devices on-site).

What's in it for you

• £98,000 - £110,000 annual salary
• 12-month contract - inside IR35 - PAYE
• Fully on-site in Cambridgeshire - hardware-heavy environment
• Proper engineering role: ownership, impact, and technical depth

What you'll do

• Build and maintain scalable Python tooling that supports ML research workflows
• Improve reliability: fix bugs, optimise performance, and ship iterative enhancements in sprints
• Integrate with distributed compute, data storage, and internal development systems
• Work closely with researchers as your "users" - gather requirements, translate ambiguity into working software
• Write clean, maintainable code and debug problems that span systems

Must-haves

• 4+ years commercial experience building software in Python
• Hands-on with PyTorch or TensorFlow in production or applied ML environments
• Experience with distributed or high-performance compute
• Comfortable working with large/complex ML datasets and research-style workflows
• Good communicator: you can explain what you're building and why, not just ship code

Nice to have

• Audio / DSP-adjacent ML exposure
• ML tooling/pipelines built for researchers or data scientists
• Linux/Windows scripting
• CI/CD and deployment experience in complex environments

How your day looks

You'll spend time building new features into internal tools, responding to researcher requests, fixing what's broken, improving what's slow, and collaborating with the infra team in a sprint cadence. This is hands‑on engineering with real ownership.

How to apply?

If you're a Python engineer who enjoys solving real systems problems for ML teams, share your CV and we'll talk through fit, on‑site expectations, and interview process.