Senior/Lead Machine Learning Engineer

present of AI; Greater London; Hybrid; GBP 90,000 - 140,000; Full time

Job summary

Mimica turns employee desktop activity into process maps and AI agents that execute the work, per the company.

The senior/lead ML engineer will own end-to-end ML pipelines, from data and model work through evaluation and production rollout, ensuring models read messy real-world work and perform reliably in production.

Qualifications

• Experience building and shipping ML systems in production.
• Strong judgment on experiments and model selection.
• Ability to manage accuracy, cost and reliability in production.

Responsibilities

• Own ML pipelines end-to-end from data work to production rollout.
• Design and evaluate LLM judges and evaluation frameworks.
• Balance accuracy, cost and latency in deployed models.

Skills

ML engineering

Production ML

Experimentation

Cost & latency tradeoffs

Tools

LLM evaluation

Job description

A senior or lead machine learning engineer is expected to own ML pipelines end to end, from data and model work through evaluation and production rollout. A strong candidate brings solid experience building and shipping ML systems, good judgment about experiments, and care for accuracy, cost and reliability. At a company turning desktop activity into process maps and AI agents, that means making models that read messy real world work and hold up in production.

How it fits the AI era

This role sits close to the AI shift, since the product uses models to understand work and act on it. Much of the craft is now about evaluation, LLM judges and turning promising experiments into dependable systems. What stays human is judgment about what good looks like, which tradeoffs matter, and how to keep results trustworthy.

How it builds AI experience

The work offers hands on practice with LLM evaluation, judge design and experiment driven improvement of production pipelines. Building that experience means running careful comparisons, writing clear evals, and learning where models fail on real activity data. Experience with cost and latency tradeoffs in deployed ML systems helps.

Worth showing

• End to end ML ownership
• LLM evaluation and judges
• Accuracy and cost tradeoffs

This is our own read, written from the employer's posting and the company's Spotlight. The employer's posting is the source of truth for the role.

About Mimica

Mimica turns employee desktop activity into process maps and AI agents that execute the work, per the company.