DevOps+MLOps+PythonML

Infosys Limited is hiring for its Data & Analytics Unit in Bangalore. The role combines DevOps and MLOps to automate delivery of machine-learning services and help data science teams run Python ML workloads in production.

Responsibilities

• Design and maintain CI/CD workflows for building, testing, releasing and deploying ML and supporting services. Automate infrastructure and configuration to keep development, staging and production environments consistent.
• Build reproducible, traceable pipelines for model training, validation, packaging and deployment. Manage model versions and artifacts, support controlled rollouts such as canary or blue-green deployments, and establish performance monitoring, drift signals and feedback loops.
• Improve reliability through monitoring, alerting, incident response and post-incident changes. Work with data scientists on testing, packaging and runtime optimization of Python ML code; define logging, metrics and SLO standards, maintain documentation and runbooks, and participate in code reviews. Recommend improvements to security, scalability and cost efficiency.

Required qualifications

• Bachelor of Engineering, BTech, MTech, MSc or MCA, or equivalent practical experience.
• 2–3 years of hands-on experience in DevOps and/or MLOps-focused engineering roles; working knowledge of CI/CD and deployment automation; practical experience supporting Python-based ML workloads, including packaging, environments, dependencies and runtime troubleshooting; and strong Linux fundamentals, networking basics and system troubleshooting.

Preferred qualifications

• End-to-end ML production experience spanning training pipelines, model registries or artifacts, deployment and monitoring. Exposure to Docker and Kubernetes; Terraform or Ansible; MLflow or Kubeflow; and Airflow. Hands-on exposure to deploying LLM-enabled applications, inference optimization and evaluation or monitoring approaches is also preferred, along with strong communication across engineering and data science teams.