MLOps+ AI Engineer

About the Role

Build and operate the MLOps pipelines that take AI/ML and GenAI models from experimentation to production: packaging, CI/CD delivery, model serving and monitoring. A hands-on engineering role bridging data science and enterprise deployment.

Experience: 3-7 years in ML/AI engineering or DevOps

Location: Bangalore (Whitefield) - Work from office, 5 days a week

Notice Period: Immediate joiners preferred (0-15 days)

Interview Mode: In-person interviews (Whitefield, Bangalore)

Education: B.E./B.Tech in CS, Software Engineering or Data Science

Please note: An online assessment is mandatory as part of the selection process.

Key Responsibilities

• Build end-to-end pipelines - data ingestion, training, evaluation, packaging, versioning and deployment
• Develop and maintain Jenkins CI/CD for Dev to QA to Production promotion with multi-stage gates
• Deploy model-serving APIs on AKS using FastAPI and vLLM; apply ONNX/TensorRT optimizations
• Set up observability - drift detection (Evidently AI), Prometheus/Grafana, Azure Monitor
• Apply DevSecOps practices - Key Vault, Managed Identity, SonarQube, Trivy/Snyk
• Build REST and WebSocket applications using FastAPI

Must-Have Skills

• 3+ years in ML/AI engineering or DevOps with hands-on production MLOps pipeline experience
• CI/CD tooling: UV, Ruff, Pyrefly, Jenkins (strong), Azure DevOps, GitOps concepts; Git and pre-commit workflows
• Databricks ML pipelines (Delta Lake, Workflows, MLflow), Asset Bundles and PySpark
• Model serving: FastAPI, Docker, AKS; exposure to vLLM and ONNX/TensorRT optimization
• Python (strong - FastAPI, Pydantic, async), Bash, YAML/SQL scripting
• Cloud knowledge - Azure/AWS/GCP storage and AI related services
• Databases and storage: PostgreSQL, Redis, ADLS Gen2
• Understanding of containerization, Helm and infrastructure automation

Nice to Have

• Airflow, DVC and experiment tracking (W&B / Comet ML)
• Terraform, KEDA, Azure APIM and AAD RBAC configuration
• LLM fine-tuning pipelines; Ray Serve or BentoML exposure
• Groovy (Jenkinsfile)
• Manufacturing or semiconductor domain experience