End-to-End ML Scientist — Growth & Experimentation

Metica; England; On-site; GBP 90,000 - 130,000; Full time

Job summary

Metica is scaling its AI-native growth platform for mobile games studios, unifying user acquisition and in-game monetisation. The role leads ML systems from concept to production, owning training, deployment, and monitoring while tying model performance to business metrics.

You'll design rigorous experiments and collaborate with engineering, product and commercial teams to drive impact. You will apply advanced inference techniques, build data pipelines, and own releases with accountability for

Qualifications

• Expert Python and SQL with track record in production ML systems and data pipelines.
• Strong grasp of experimentation and statistics, and uncertainty optimization.
• Data-driven and able to collaborate across engineering, product and commercial teams.
• Ownership of production business metrics beyond research models is a plus.

Responsibilities

• Own ML systems end to end from hypothesis through deployment and monitoring.
• Write production-quality training and inference code and manage releases.
• Design and run A/B and multivariate experiments with power analysis and significance checks.
• Apply contextual bandits, causal inference, uplift modelling, and off-policy evaluation.
• Analyze large-scale datasets and translate findings into actionable recommendations.

Skills

Python

SQL

Production ML

Data pipelines

Experimentation

Tools

PyTorch

XGBoost

Spark

Apache Iceberg

Ray

MLflow

SageMaker

Airflow

Job description

Metica is scaling its AI-native growth platform for mobile games studios, unifying user acquisition and in-game monetisation. The role leads ML systems from concept to production, owning training, deployment, and monitoring while tying model performance to business metrics.

You'll design rigorous experiments and collaborate with engineering, product and commercial teams to drive impact. You will apply advanced inference techniques, build data pipelines, and own releases with accountability for