← Back to projects
Eurowings · Lufthansa Group•6-month platform build

Enterprise MLOps Platform

Infrastructure for operating ML at scale

Built the platform layer for governed ML deployment on Azure Databricks, standardizing CI/CD, registry workflows, monitoring, and access control across teams.

MLOpsDatabricksMLflowUnity CatalogAzure DevOps

Business Problem

Teams deployed models inconsistently, with weak governance, fragmented monitoring, and no reproducible delivery path.

Architecture

  • Git and Azure DevOps for versioned pipelines
  • Databricks MLflow tracking and model registry
  • Unity Catalog and Feature Store for governance
  • Serving endpoints and monitoring dashboards for production models

Technical Decisions

  • Standardized around Databricks and MLflow to reduce platform sprawl and make delivery repeatable.
  • Used Unity Catalog and Key Vault integration to align access control with enterprise governance.
  • Built reusable CI/CD templates rather than one-off project pipelines to reduce setup time across teams.

Production Challenges

  • Solved secret management with Azure Key Vault integration.
  • Implemented drift, latency, failure, and feature-quality monitoring for deployed models.
  • Created reusable deployment templates to move teams from ad hoc releases to governed delivery.

Outcomes

Reduced deployment time from weeks to hours
Standardized delivery across ML teams
Improved reproducibility and platform governance