Real-world projects with measurable outcomes. Each case study details the challenge, approach, and quantified results.
10 case studies · 37 tags · 28,179 words
Filter by tag
How we replaced Datadog and Splunk with a self-hosted LGTM stack, moved more than 100 engineering teams, cut annual spend by over $13 million, and decided which operational platforms were worth owning.
Read case studyTwo providers, one LiteLLM gateway, and a lock-in budget we actually drill. How teams get onto Bedrock without a provider SDK, which workloads pay for the hybrid surface, and the three-layer observability that keeps both vendors auditable.
Read case studyReplacing end-of-life ingress-nginx with Traefik across 30 Kubernetes clusters and 120 namespaces with zero outage minutes. Parallel-run weights, a DNS flip on ingressClassName, Middleware CRs for the annotation debt, and an AI skill wired to the mapping from discovery.
Read case studyTaking a Medicaid and ACA enrollment platform from baseline to 15x load without customer-visible incidents by fixing compute, streaming, database, and API bottlenecks during the November spike.
Read case studyArchitecting a geographically isolated, GDPR-compliant, active-active platform across 5 AWS regions with 99.999% availability.
Read case studyThree weeks to instrument an LLM system nobody could debug.
Read case studyWhen US East explodes at 3am, traffic shifts automatically. Customers see a blip. You stay asleep.
Read case studyMigrating 300+ enterprise applications from on-premises Kubernetes to Azure without a single customer facing incident.
Read case studyA six-dimension signal contract for Platform, SRE, Dev, and DevOps: what to instrument, who owns it, how to alert, and how to triage when dashboards lie.
Read case studyHow we evolved a fintech API from 800K to 10M+ daily requests through architectural decomposition, eliminating the monolith bottleneck while maintaining zero downtime and cutting P95 latency by 58%.
Read case study