MLOps Engineer

About the role

The Technology Innovation Institute is building a startup to industrialise a secure, AI-enabled decision-support platform that fuses multi-sensor data such as geospatial layers, imagery and video. As an MLOps Engineer you will own the reliability and scalability of ML and LLM services, creating end-to-end pipelines, deployment automation and observability in a fast-moving environment.

Key responsibilities

• Design, implement and operate full ML/LLM delivery pipelines from data ingestion to model training, fine-tuning, evaluation, packaging and deployment.
• Build CI/CD workflows for models and services, including automated testing, validation gates and rollback mechanisms.
• Standardise experiment tracking, model and version lineage, and artifact management for datasets, prompts, checkpoints and embeddings.
• Implement monitoring and observability covering latency, cost, data drift, quality signals and safety guardrails.
• Optimise inference performance and cost through batching, caching, quantisation and hardware selection.
• Define and enforce environment and dependency management across development, staging and production.
• Collaborate with platform engineers, security teams and product owners to ensure scalable serving patterns, access controls and release readiness.

Required profile

• At least 5 years of experience in MLOps, DevOps or data platform roles with production deployments of ML or LLM systems.
• Proven ability to work in fast-paced product or startup environments.
• Strong problem-solving skills and the capacity to unblock teams quickly.

Required skills

• ML lifecycle tools: MLflow, Weights & Biases.
• Model serving: FastAPI, Triton, Ray Serve.
• Orchestration: Airflow, Dagster.
• Observability: OpenTelemetry, ELK, Prometheus, Grafana.
• Cloud platforms: GCP, Azure, AWS.