Principal MLOps Engineer

Why PayNet / Why Now

• National payments infrastructure with real economic and systemic impact
• Organisation entering a phase of greater scale, scrutiny, and performance expectations
• Leadership demanding clearer differentiation, stronger governance, and better data
• People function expected to shape outcomes, not just run processes

TL;DR

• Own production‑grade Machine Learning Operations (MLOps) platforms powering fraud and risk intelligence
• Decide how Machine Learning (ML) models are promoted, rolled back, and governed in production
• Build secure, auditable platforms across Amazon Web Services (AWS) and hybrid environments
• Partner with Data Scientists to turn models into reliable, explainable scoring services

Why This Role Matters

• Fraud models only create value when they are stable, explainable, and production‑ready
• This role governs the boundary between ML innovation and real‑world financial impact
• Engineering decisions here directly affect system resilience and regulatory confidence
• You enable PayNet to scale Artificial Intelligence (AI) without compromising trust

What You Will Actually Do

• Own end‑to‑end MLOps productionisation for fraud and risk intelligence use cases
• Build and operate Continuous Integration / Continuous Deployment (CI/CD) pipelines for model and service release
• Design and enforce model lifecycle management, including versioning, retraining, and redeployment
• Architect and operate secure AWS and on‑premises hybrid infrastructure for ML platforms
• Implement platform standards using Infrastructure as Code (IaC), containerisation, and Identity and Access Management (IAM)
• Ensure deployments meet audit, security, and regulatory requirements without sacrificing stability

Examples of This Role in Practice

• Decide whether a fraud model can be safely promoted during elevated transaction risk
• Design rollback mechanisms when a real‑time scoring service degrades latency
• Convert experimental notebooks into governed, auditable production pipelines
• Balance model accuracy, infrastructure cost, and response time at national scale

What Will Help You Succeed

• Experience building and operating production ML systems, MLOps platforms, or large‑scale DevOps environments
• Strong proficiency in Python for ML pipelines, model packaging, automation, and service integration
• Deep hands‑on expertise with AWS architecture, including secure networking and high‑availability design
• Proven ability to design CI/CD pipelines for ML services with gated releases and controlled promotion
• Experience with IaC tools such as Terraform and container orchestration using Kubernetes and Helm
• Familiarity with distributed workloads (e.g. Apache Spark or Ray) and orchestration tools such as Apache Airflow or Prefect