ML Platform Engineer at SR2 - Socially Responsible Recruitment, London, 6 Months, to £575 per day

Senior ML Platform/MLOps Engineer

Remote | Outside IR35 | 6 months | Up to £575 per day
Start: August
Industry: Global payments

We are looking for a Senior ML Platform/MLOps Engineer to join the ML platform team within a global payments business.

The team builds the Real Time machine learning and feature infrastructure behind fraud detection, scoring transactions at scale within strict latency requirements. This is a hands-on platform engineering role, not a data science position, with Databricks sitting at the heart of the training, feature and model life cycle.

• Building and evolving ML platform infrastructure across Databricks and AWS
• Developing scalable batch and streaming pipelines using PySpark and Spark Structured Streaming
• Improving Delta Lake architecture, data reliability and online/offline feature parity
• Working with Unity Catalog to support governance, access control and secure data management
• Building and improving MLflow-based model tracking, retraining and deployment workflows
• Supporting feature-store infrastructure across Databricks, DynamoDB and Redis
• Developing Real Time features using Kafka and stateful streaming patterns
• Improving model serving, canary releases, shadow deployments and champion/challenger testing
• Optimising slow and costly training pipelines
• Managing AWS infrastructure through Terraform
• Improving observability, security, reliability and multi-region resilience

• Strong Python and production ML platform or MLOps engineering experience
• Deep hands-on Databricks experience across PySpark, Delta Lake, Unity Catalog and MLflow
• Strong Spark Structured Streaming and Kafka experience
• Deep AWS experience, ideally across ECS, EC2, DynamoDB, IAM and KMS
• Strong Terraform and infrastructure-as-code experience
• Experience deploying and serving models using tools such as FastAPI, Seldon MLServer or NVIDIA Triton
• Strong understanding of CI/CD, monitoring and safe model rollout strategies
• Previous consulting or client-facing delivery experience
• Confidence owning workstreams independently in environments with strict latency, reliability and cost constraints
• Experience with Redis, feature stores, multi-region AWS, payments, fraud or other low-latency financial services environments would be a strong advantage.