Staff ML Engineer, North Modelling Evals

Neura Market; Greater London; Hybrid; GBP 90,000 - 130,000; Full time

Benefits offered by this job

Lunch stipend

Health and dental benefits

RRSP matching / Pension

Parental Leave top-up

Enrichment benefits

6 weeks paid vacation

Home office stipend

Job summary

Cohere is seeking a technically exceptional individual for the North Modelling (Evals) team. You will own the evaluation strategy, align metrics with enterprise workflows, and drive product outcomes through careful measurement and feedback loops.

You will work with North product teams, customer-facing teams, and central modelling groups to ensure model progress translates into real-world value. This role blends applied MLE rigor with product impact, demanding hands-on creation of evals, data

Qualifications

• Experience building evals for LLM-powered or AI-product systems.
• Strong focus on measurement, data quality and regression tracking.
• Ability to translate product needs into evaluative signals and metrics.

Responsibilities

• Own the eval strategy for North: define measurements across agent workflows and enterprise use cases.
• Build evals from user feedback, production data, and privacy-preserving logs to guide product decisions.
• Create systems that turn learnings into actionable eval improvements for model updates.
• Extract insights from eval results and translate into recommendations for central modelling teams.

Skills

Eval strategy

Model evaluation

Production ML

Collaboration

Tools

Python

ML tooling

Job description

Cohere is seeking a technically exceptional individual for the North Modelling (Evals) team. You will own the evaluation strategy, align metrics with enterprise workflows, and drive product outcomes through careful measurement and feedback loops.

You will work with North product teams, customer-facing teams, and central modelling groups to ensure model progress translates into real-world value. This role blends applied MLE rigor with product impact, demanding hands-on creation of evals, data