Machine Learning Engineer

Amberes Recruitment; Greater London; On-site; GBP 85,000 - 110,000; Full time

Benefits offered by this job

Bonus

Benefits

Equity

Job summary

Amberes Recruitment seeks a hands-on AI/ML engineer to take large, complex models from research into production. You will deploy models that run in real customer environments with limited compute, working closely with the software team to shape the ML stack.

We value 3+ years building AI systems for real business problems, C++ inference experience (ONNX Runtime, TensorRT, libtorch), and strong DL knowledge (CNNs, time-series, transfer learning).

Qualifications

• 3+ years building and deploying AI systems that solved real business problems.
• Experience deploying ML inference in C++ (ONNX Runtime, TensorRT, libtorch).
• Strong deep learning knowledge, including CNNs, time-series models and transfer learning.
• A degree in computer science, engineering or a related field, or equivalent experience.
• A practical, problem-solving mindset and right to work in the UK.

Responsibilities

• Build and deploy data and inference pipelines handling many data types (images, time-series, point clouds).
• Deploy complex models to production where standard tools fall short.
• Collaborate with software engineers to optimize ML stack and edge deployment.

Skills

C++ deployment

Deep learning

Time-series models

Transfer learning

CUDA

GPU programming

Education

Bachelor's degree in CS/Engineering or equivalent

Tools

ONNX Runtime

TensorRT

libtorch

Job description

• Salary: £90K + Bonus + Benefits + Meaningul Equity

The Business

• This is a pre-Series A, seed-backed AI company working at the cutting edge of robotics. It was spun out of Imperial College London by a group of top PhD engineers, and in under two years it has raised £5m and won paying customers across aerospace, defence and medical devices.
• The platform takes data from many different sensors on industrial machines and turns it into real-time insights that help manufacturers produce safety-critical parts with confidence. The team is currently circa 10 people and growing.

The Position:

• You'll take large, complex AI models out of research and into production, making them fast, reliable and easy to maintain. The models you deploy will run in real customer environments, often with limited computing power, so smart engineering really matters here. You'll work closely with the software team and have a big say in how the ML stack is built.

What You'll Work On

• Data and inference pipelines: Build and deploy pipelines that handle many types of data, including images, time-series and point clouds.
• Deploying complex models: Find new ways to get large, demanding models running well in production, even where standard tools fall short.
• Performance at the edge: Partner with the software engineers to solve tricky deployment problems on hardware with tight resource limits.

Ideal Candidate Profile:

Someone who loves getting AI working in the real world, not just in a notebook, and who can move quickly without cutting corners.

You'll need:

• 3+ years building and deploying AI systems that solved real business problems
• Experience deploying ML inference in C++ (e.g. ONNX Runtime, TensorRT, libtorch)
• Strong deep learning knowledge, including CNNs, time-series models and transfer learning
• A degree in computer science, engineering or a related field, or equivalent experience
• A practical, problem-solving mindset
• The right to work in the UK
• Deploying AI in regulated or safety‑critical industries such as aerospace, medical devices or automotive
• GPU programming with CUDA
• Papers at top AI conferences such as NeurIPS, ICML, ICLR or CVPR
• A background in robotics, industrial sensing or signal processing
• Startup experience, or a track record of thriving in fast-moving, uncertain settings

Why Join

• Early seat at a well-funded Imperial spinout that already has paying customers
• Work on hard, unusual deployment problems you won't find in most ML roles
• Real ownership of the ML stack in a small, expert team
• Your work ends up in safety-critical products used by major manufacturers

Applicants' Note:

An Amberes job posting can reach over 1,000 applicants, and we advertise across dozens of openings at any one time. The team does its absolute best to acknowledge every application in good time and reply with updates. However, because of the volume, we are not always able to reply to everyone, and we apologise if that happens to you. Applications are received and held via LinkedIn. We do not store applicant data on our own systems without your consent.