Till Wiese

Projects

Platform engineering and SRE

• I led the migration of our kube-prometheus-stack monitoring to Google Managed Prometheus on all clusters, converting the alerting rules one subsystem at a time and retuning them to stop false alerts.

• I developed a Python script to translate the module dependencies of a monorepo covering ~128 GCP projects plus AWS, Azure, Cloudflare and GitLab, applied through Terragrunt and Atlantis, into a format Atlantis could parse.

• In 2026 I was part of the team migrating a sister company’s infrastructure from on-premises to GCP, building the GKE clusters, Cloud SQL, Memorystore Redis and Private Service Connect networking in the process.

• I introduced our developer teams to Claude Code with MCP servers for GKE, Cloud Logging, Cloud Monitoring, Cloudflare and GitLab to run cost audits and root-cause analyses.

Cloud cost

• I audited the cost of our GKE clusters and identified savings of ~25% of the monthly run-rate.

• I halved the cost of our metrics and monitoring stack by optimizing the Prometheus metrics scraping, the rules and the alerting configuration.

• I built a reusable Terraform budget alert module with current and forecasted spend thresholds for GCP and AWS.

Security and edge

• I defended our shops against a distributed crawler campaign that sent tens of millions of faceted-search requests by identifying common traffic patterns and mitigating them sustainably.

• I decreased our network costs by introducing heuristic Cloud Armor rules to our GKE clusters, thusly blocking unwanted traffic before it even reached our shops.

Projects

• I led the technical evaluation and subsequent decision of the GitOps approach via Argo CD and the underlying technical details.

• I developed an Ansible wrapper playbook to adjust OpenShift “Installer Provisioned Installations” to the unique Deutsche Börse Cloud environment.

• I was a key project member for the migration of the Deutsche Börse CMS serving most Deutsche Börse web sites from an on-premise to a container-native architecture.

Projects

• I led the effort to automate the deployment and configuration management for web applications by using Ansible for config rollouts and Jenkins for application deployments.

• I developed a dynamic Docker Swarm monitoring tool in Python 3 to enable the company internal monitoring solution to properly monitor dynamically changing Docker Swarms.

• I was a key player in the effort to transform existing web applications for cloud platform (AWS and Microsoft Azure) viability via Terraform and Ansible, thusly saving costs by automatically scaling applications horizontally.

Projects

• I was part of a team to plan and establish automated configuration management on >2000 SUSE Linux Enterprise servers spanning two data centers. We used our own created tool written in Perl and Bash and set up our own failover PostgreSQL method for it to cover the characteristics of the underlying environment.

• I contributed to significantly reducing the ‘days to data center’ for new server orders by semi-automating the ordering and ESX installation processes.

Projects

• I implemented a roaming user profile management system for Windows on an openSUSE server using OpenLDAP and Kerberos.