From crisis to resilience: stabilizing a 6M-transaction-a-day platform

The situation

Swiss Automotive Group is Europe's largest independent auto-parts distributor — 1.8 million SKUs, market leader across 12 EU markets, and over 95% of transactions now running through its e-commerce platform. As volume scaled past 6 million transactions a day, the legacy middleware supporting SAG Cloud collapsed mid-trading. The Kubernetes layer was in place, but everything around it had drifted far from modern standards — a single-point-of-failure stack that had become a security and availability liability.

What we did

• Stabilized production under active failure conditions — restoring trading in days, not weeks.
• Built a brand-new cloud-native infrastructure in parallel, following Azure Cloud Adoption, Well-Architected, and Zero Trust frameworks.
• Introduced fully auditable GitOps release cycles, autoscaling, and Infrastructure-as-Code via Terraform.
• Swapped ad-hoc auth and secrets for OIDC passwordless service-to-service, Bitwarden for clients, Azure Key Vault for apps, and an XDR solution for detection.
• Migrated off the legacy environment in weeks with no customer-visible disruption.
• Drove a FinOps practice with the internal team — reserved instances, dev/test pricing, log-volume optimization.

The outcome

• Zero hours of downtime across three-plus years since cutover.
• Platform capacity expanded 50–60% with headroom for peak-traffic and disaster recovery.
• Infrastructure cost down ~60% through autoscaling, reserved capacity, and FinOps discipline.
• Log volume cut 60%; log storage cost cut 40%.
• Fully automated SSL lifecycle via Let's Encrypt — certificates stop being a source of incidents.
• Security posture aligned with Azure governance policies and enterprise-grade frameworks.

Stack