Constantin Michael Weisser

The client was a B2B logistics enterprise with a stable customer base, but competition was catching up. The leaders wanted to explore ways to secure the market position and identified high lead time and failure rate as the primary issues. They faced technical difficulties such as an aging on-premises stack, complex monoliths, and scattered configuration. They were confronted with cultural hurdles, such as unclear responsibilities for incident response.

I was part of the DevOps transformation leadership team. I guided the CICD team in building CD as an in-house product. I designed the company's tech stack with the infrastructure team and advised the management team in picking key technologies. I also supported the management team in choosing KPIs that determined the transformation's success.

During my engagement, the first microservices were deployed to the infrastructure on Azure. Terraform automatically provisioned the cloud environment, which could significantly improve the infrastructure's reliability. Instead of scheduling massive releases and high risk, engineers deployed changes within hours or days of picking up a task. The management team now had access to metrics such as Lead Time for Changes and Deployment Frequency.

I've contributed to this project as a DevOps engineer. I initially developed an automated cloud architecture to run around twenty microservices on Azure. Software from six product teams was automatically built and deployed via Azure DevOps pipelines. The deployment target was AKS on the Azure cloud. The whole infrastructure was described and automated via Terraform modules. One of the biggest challenges was the unavailability of operations engineers that could help to operate the infrastructure. Consequently, I standardized systematic manual development processes and provided reusable solutions to engineers to speed up rollouts and diminish cognitive load.

This client hosted dozens of similar environments, one per customer. Initially, all environments were traditional on-premises deployments that engineers built and maintained manually. The customer wanted to reduce the set up time from weeks to hours. They were looking for a scalable solution to add a new customer on demand.

I advised my client to select ideal technologies for cloud-based deployment on Azure. The client had only a few people for operations, so I chose scalable low-maintenance solutions such as Azure Appservice. I built Terraform modules as the blueprint for a single environment. Automation and configuration management complemented the code to quickly scale up to dozens or even hundreds of instances. I devised an update strategy and best practices for monitoring and log management at scale. Since the application had only been used on-premises before, I supported the engineers in turning it into cloud-native services and debugging tricky platform dependencies.

With my solution, my client could build a new environment in minutes. All deployments were independently configurable in size and feature flags. That way, they could adjust the deployment to the exact needs while keeping management overhead to a minimum.

This project's goal was to extend an in-house contract testing platform that ensures safe synchronous and asynchronous communication between more than a thousand microservices. The platform enables independent and concurrent development of microservices to step up the company's product development.

I contributed as an engineer implementing support for multi-country deployments. Such deployments cause regional differences in the microservices' communication patterns. The lack of support put the company's multinational strategy at risk. I led feature development, devised a migration strategy, and ensured a frictionless rollout without interrupting the hundreds of daily test runs on which engineers rely. I also automated metrics collection to follow up on feature adoption.

The added feature support enabled the company to reuse products in a multinational context. Successful ventures can be transferred more easily into another country, increasing overall business agility. The platform delivers tens of thousands of pipeline checks and almost four thousand pre-deployment checks to this day while keeping communication errors in production to an absolute minimum.

A Spring Boot and Kubernetes-based cloud-native SaaS application that provides well-structured and auditable communication between construction site managers, lead workers, and subcontractors on large-scale construction sites.

I contributed to this project as a back-end software engineer. I worked closely with the product owners and other stakeholders to deliver new features, which are digitized versions of the usual civil construction modus operandi. I provided vital pieces to cross-cutting solutions, such as solving the exactly-once semantics for sending Apache Kafka messages and committing to a database.

I was also responsible for CI/CD infrastructure built on Jenkins with automatically provisioned worker nodes on AWS.

A client from the financial sector accumulated several unconnected databases with customer records over the years. Back-office clerks had to consult all those systems manually when trying to identify a customer and the products they already had.

As a back-end software engineer, I designed and implemented a central customer identification system. Data was regularly pulled from the existing customer databases and potentially merged with records from other data sources. The merge criteria were built as an extensible rule set. I created a straightforward user interface that enabled the search and visualization of customer records. This central entry point reduced clerks' effort from more than half an hour to less than a minute for each customer request.