Senior MLOps Engineer

1. ML Deployment Architecture & CI/CD

• Lead the design and build of CI/CD pipelines for ML model packaging, testing, versioning, and deployment across dev, test, and production environments.
• Define and enforce deployment standards — staged rollouts, canary releases, rollback procedures — as reusable patterns for the team.
• Architect containerised model serving solutions, making considered decisions on orchestration, resource allocation, and environment parity.
• Own the reusable deployment framework: templates, shared libraries, and pipeline components that reduce rework across model teams.

2. Monitoring, Observability & Production Reliability

• Design and implement end-to-end observability for model performance, data drift, pipeline health, and inference quality in production.
• Set monitoring coverage standards and drive adoption across all deployed models — not just newly deployed ones.
• Lead root-cause investigation for complex production incidents; own post-incident reviews and structural fixes.
• Ensure all deployed models have defined SLOs, alerting thresholds, and clear incident ownership before going live.

3. Platform & Infrastructure Design

• Own training, retraining, and batch-scoring pipeline architecture — including scheduling, dependency management, and failure-recovery design.
• Make technical decisions on model registry structure, artifact lineage, and versioning conventions across the team.
• Lead environment architecture decisions: base images, dependency isolation, environment parity between training and serving.
• Partner with platform and architecture teams to optimise compute resource usage and infrastructure costs for ML workloads.

4. Technical Leadership & Enablement

• Provide technical direction and design review for other ML Ops engineers on the team.
• Define and document deployment patterns, runbooks, and operational standards that enable team self-service.
• Work with Data Scientists to translate model requirements into production-ready, maintainable deployment designs.
• Partner with Data Engineering and Governance teams to ensure data feeds, lineage, and compliance requirements are met in production.
• Communicate technical decisions, risks, and trade-offs clearly to cross-functional stakeholders and leadership.