Customer-Portal

• Agile, valuable and future-proof: further development, modularization and technical modernization of the personal customer portal My AXA using an agile Scrum approach.

• Digital transformation: integrating new REST APIs to expand the digital product and service portfolio in coordination with internal and external partners

• Monolith modularity: splitting a complex monolithic application along clear business and technical responsibilities

• Clean code: continuously refactoring complex code for readability, maintainability and test coverage (pathfinder principle)

• Agile learning: actively and continuously presenting the latest developments and innovative technologies/concepts in a weekly cross-team Know-How Transfer meeting

• Speaker coordination: recommending and presenting an external speaker for domain-driven design and modularization

• Consumer-driven: advising and supporting other teams in designing interfaces to connect and extend My AXA functionality

• Integration testing: introducing WireMock to simulate backend API calls in automated and manual developer tests

• Standardization: replacing manual caching and resilience solutions with industry standards like Spring Cache (@Cacheable) and Resilience4J (@CircuitBreaker, @Bulkhead, @TimeLimiter, etc.)

• Data validation: uncovering and fixing errors in existing request validation using JSR-170 Java Bean Validation, coordinating with clients and developing a “silent bean validation” simulation approach

• Code generation: improving and cleaning up the code generator for transforming OpenAPI contracts into client libraries

• Deep dive: analyzing Lombok‐generated bytecode output and fixing errors

• Observability: developing dashboards and log analyses with Loki and Grafana Cloud

• Operations: daily monitoring and reporting of production and QA systems

• Backend Developer, DevOps Engineer

• Java 21, Spring Boot 3, Spring Integration, Spring Cache, REST, Java Bean Validation, Resilience4J, Mustache, Freemarker, Lombok, JAX-WS, Mockito, Docker, OpenAPI, Swagger, WireMock, Loki, Grafana Cloud, Graylog, Jenkins, Tekton, Rundeck, SonarQube, Checkstyle, Spotbugs, Nexus, Red Hat OpenShift, GitLab, GitHub, IntelliJ IDEA 2025, Maven, Gradle, Jira, Confluence, Scrum

Press Center

• Distributed system providing weekly program information and media content
• Refactored an existing distributed application landscape to approach microservice granularity and a dynamic, scalable runtime architecture with Spring Cloud and Docker Swarm. Developed a new public REST service for weekly program, VOD and EPG metadata and media content for network operators and external platform partners. Implemented multi-layer security for authentication header validation and authorization against various company databases using path matching (regex) and request filtering as custom Spring Boot starter modules loaded via classpath dependencies with autoconfiguration. Ensured balancing and availability of about 100 application processes using Spring Cloud’s Eureka, Zuul and Ribbon and deployed successfully. Later replaced Zuul, Eureka and Config Server with Docker-only solutions. Additional tasks included migrating standalone Tomcats to Spring Boot and extending Swing and JavaFX clients.

Content Information Service

• Supplying network operators with metadata on linear and non-linear programs

• Multiple apps in the media group manage metadata for all channels’ linear and non-linear content. With continuous expansion of the video-on-demand area, requirements have grown more complex for internal systems and external operators. The Content Information Service collects and consolidates metadata from various sources into a distributed, scalable greenfield system via a messaging interface, normalizes data using an entity-attribute-value pattern and exposes REST resources in various representations (Accept header). Domain-driven (DDD) decomposition into bounded contexts creates a highly scalable, fault-tolerant communication landscape of HATEOAS (REST) connected resources in a Docker stack.

• Network operators, channel providers and other external platform partners

• CBC and RTL Group employees

• Software Developer, Domain Architect, DevOps Engineer

• Agile process evangelist through active coaching and QA

• Microservices, Domain-Driven Design (DDD), REST, HATEOAS, Spring Boot, Spring Data (JPA), Spring Security, Spring Test, Spring EL, Spring Cloud, Netflix OSS, Zuul, Eureka, Ribbon, Hystrix, Docker (Swarm Mode), Apache Kafka, RabbitMQ, Jenkins, Nexus, Microsoft SQL Server, Hibernate, MockMVC, Swing, JavaFX, JSON, JsonPath, XML, Jira, Confluence, Scrum, IntelliJ IDEA, “make .jar, not .war”

• New customer outreach via Infopost – address rental

• Rebuilt a multi-layer Java EE application to provide address data for acquiring new customers. The system covers the end-to-end process from customer selection to lettershop delivery. To handle large order volumes economically, backoffice processing times must be reduced to a few minutes, requiring a high-performance, low-effort frontend with an interactive map for area and radial selection. In addition to HTML interaction, the system offers interfaces for document upload/download and automated sending of generated Word/PDF documents by email. The runtime scenario includes multiple Java processes (Tomcat, JBoss) persisting to an Oracle RAC via JDBC and communicating via REST/JSON with subsystems for map visuals, office documents, in-memory address counting engines, and auto-completion of area inputs.

• Deutsche Post Direkt GmbH customers

• Backoffice staff and mobile salesforce

• DPAG partner lettershops

• Software Developer, Domain Architect

• VAADIN, Spring, REST (RESEasy, Jersey), JSON, XML, JEE, JBoss, JPA/Hibernate, Liquibase, Oracle & H2, Maven, Vagrant, Jenkins, IntelliJ IDEA, Subversion

ONEXSYS

• ONEXSYS (Online Exhibitioning System) is a self-designed and developed information system for planning, organizing and running the Filtech Exhibition, the worldwide hub of the filtration and separation industry. It also serves as a content system to generate dynamic website content and handles registration—including secure payment data via HTTPS—for visitors and course participants. In 2012, after eight years of production use, a full redesign and expansion were carried out in close coordination with the client.

NOX (Next ONEXSYS)

• The next generation ONEXSYS. Alongside a fully modernized development and operations platform, NOX integrates many new business functions from the internal legacy system (ERS). All core processes and webservice functions are consolidated into a single, maintenance-friendly system. Modern, flexible technologies are key to efficient maintainability, evolution and user experience. New features can be added as service bundles at runtime and the new persistence layer with built-in search answers queries over tens of thousands of records in milliseconds on entry-level hardware.

• The application design includes the strongly schematized business core in an H2 database and a separate content-oriented persistence model for handling growing traffic data, metadata and assets like web content, documents and media.

• Internet architecture style (REST), the Open/Closed Principle deeply embedded in development and a consistent event bus form the basis of a modern, adaptable application platform for tomorrow’s requirements.

• Enterprise process platform and content system for the website

• Initiator, lead developer, technical consultant

• ONEXSYS: Struts, Hibernate, JBoss AS, Tomcat, MySQL, Apache httpd, IntelliJ IDEA, Ant

• NOX: Vaadin, jQuery, Spring, Hibernate, JPA, Lucene, AspectJ, MS SQL Server, JSR-303 (Bean Validation), POI, Google EventBus, Sling/Felix/Karaf (OSGi), MongoDB, Jackrabbit/JCR (NoSQL), H2, Tomcat, Jetty, Apache httpd, Enterprise Architect, IntelliJ IDEA, Maven

• Internal training project “pathfinder”

• Teaching practical, market-relevant skills to new employees around object- and component-oriented Java Enterprise development. Under the motto “make .jar, not .war,” we built a self-contained Spring Boot data app with JPA persistence, REST interface and responsive frontend.

• After a detailed functional analysis, we modeled the domain as JPA entities for persistence in the popular H2 developer database accessed via Spring Data repositories. We then designed and implemented the REST interface as Spring controllers, focusing on REST not only as cohesive endpoints mapping to entities but also including pseudo-resources to express high interface semantics. Proper use of HTTP status codes, parameterizing requests and responses, and the state paradigm HATEOAS, making REST partly a client concern, rounded out the service part.

• Cutting-edge frontend technologies with Apache Wicket, Bootstrap and HTML5 gave the solid base app a polished presentation and responsive design to cover all major device classes from one codebase.

• All project phases—from analysis and design discussions to issue tracking and code management—were documented in Bitbucket’s agile hosting platform.

• In under 30 days we delivered a working application, an agile configuration management setup and a successful team!

• Haeger Consulting staff

• Software Development Coach

• Wicket, Bootstrap, HTML5, CSS3, jQuery, RESTful HTTP, Spring Boot, Spring Data, JPA/Hibernate, H2, IntelliJ IDEA, Maven, Bitbucket (wiki, issue tracking, git)

Single Source Publishing

• Designed and implemented a content information system providing reusable, internationalized content elements for brochures, web and web specials. Followed corporate SOA/REST architecture. Developed online interfaces for integration into the corporate infrastructure. Designed a content repository managing elements per ISO 639 localization and mapped REST interfaces accordingly. Built a Web 2.0 admin frontend.

• Intranet application with external service integration

• Sling/Felix/Karaf (OSGi), Jackrabbit/JCR (NoSQL), Vaadin, jQuery, Google EventBus, Jetty, SOA, REST, Maven, Enterprise Architect, IntelliJ IDEA, CentOS

TCO Calculator

• Re-engineered and extended an existing web app calculating cost of ownership for various construction vehicles vs. competitors. Switched text inputs to interactive sliders and implemented financial algorithms.

• Internet application

• Felix/Karaf (OSGi), Jetty, Vaadin, jQuery, Google EventBus, Maven, IntelliJ IDEA, Debian OS

• Full-stack Developer, Technical Consultant, Concept Author

• Subsystem for synchronous substitute item determination

• For picking online orders in a REWE market open to walk-in customers, we created a software platform guiding pickers through the store and helping them gather order items quickly. If items are unavailable during picking, substitute items must be provided. Requests go synchronously via HTTP to the Substitute service, which wraps, orchestrates and aggregates multiple requests into a usable result.

• The core component for substitutes runs on a dedicated Elasticsearch cluster using HTTP More-Like-This searches for each item. Matching substitutes get availability data and scored sorting. The “Substitute service” bounded context maintains its own data store (Elasticsearch index) for substitutes and a relational JPA/Cloud SQL model for stock info. Local stores are fed from external Kafka messages: continuous updates and full reloads by resetting topic offsets. Reindexing and stock rebuild are Jenkins jobs.

• Availability data accuracy is enhanced by synchronous REST calls that consider future availability and quantities already picked for other orders.

• With synchronous end-to-end processing and a strict response time guarantee, availability and fault tolerance are primary quality goals. Achieved via Kubernetes and circuit-breaker patterns from the Resilience library Hystrix.

• Use of meaningful HTTP status codes, hypermedia/self-describing resources (HATEOAS) and intuitive resource identifiers ensures full REST maturity (Level 3). Complete openAPI 3.0 documentation via Swagger meets high customer quality standards and architectural constraints.

• Fulfilment center pickers

• Software Developer, Architect

• Microservices, Domain-Driven Design (DDD), RESTful HTTP, HATEOAS, Spring Boot, Spring Data (JPA), Spring JDBC, Hibernate, resilience, circuit-breaker, Hystrix, Apache Kafka, Jenkins, Spock Framework, MockMVC, Groovy, WireMock, Swagger, Jackson, Gradle, git, Docker, Kubernetes, Jira, Confluence, Bitbucket, Scrum, IntelliJ IDEA

Big Data Dashboard

• Designed and implemented a web UI for visualizing stats from a big-data platform. The focus was an appealing, modern look with high usability using Bootstrap, chart components, jQuery plugins and modern UI patterns like progressive disclosure.

• Users: REWE Group management

Marktdatenbank

• Built a web info system for managing in-store placement elements. Included a JPA data layer, REST API for other systems, batch processing, an in-memory mock client lib and a web UI. Security requirements included SSL encryption and a role-based access concept. Features could be toggled per build. The app reused some existing DB2 tables and included SQL scripts to create and seed new ones.

• Users: store managers

• Full-stack Developer

• Apache Wicket, Bootstrap, HTML5, CSS3, jQuery, AJAX, Spring, Hibernate, JPA, SQL, Apache Tomcat, JBoss CDI, DB2, IntelliJ IDEA, Eclipse, DBVisualizer, Subversion, Maven, Quickbuild (CI), Jira, Confluence

• CMS-based eCommerce platform for ZEG brand shops and partner dealers

• ZEG is a purchasing cooperative of 960 bike stores in Europe and develops exclusive models in collaboration with leading manufacturers for their online platform.

• To expand payment providers, delivery/sorting options, improve availability detection, connect more dealers (from 5 to over 200), comply with new data protection laws and implement many detail improvements, the ZEG platform needs a large upgrade.

• The platform is a monolith. Despite microservice trends, well-structured, maintainable monoliths are possible. However, the system grew pragmatically under low-agile processes and fell far short in stability, flexibility, performance and maintainability, making development slow and inefficient.

• After proposing agile methods, better documentation and setting up QS and CI infrastructure, I was given the chance as co-lead to implement these improvements.

• Positive impacts on quality and team feedback acceptance led to my appointment as technical lead. I then set the core goals:

• Define clear business areas and key abstractions (domain entities)

• Drive domain formation, design clean interfaces (RESTful HTTP)

• Build independent microservices targeting modern orchestration runtimes

• Shift to reactive programming and non-blocking communication

• Externalize new cross-functional frontend on performant React

• Increase resilience via external circuit-breakers (sidecar pattern) and failover strategies

• Establish monitoring, alerting, metrics and self-healing

• Remove redundant code and obsolete features

• Raise automation, write Jenkins Job DSL scripts

• Migrate from centralized Bamboo agents to Infrastructure as Code

• Instill clean code, software quality and 12-factor compliance values in the team

• Stakeholders: internet end users, online editors and partner dealers on the ZEG platform

• Technical Lead, Software Developer, DevOps Engineer

• Java 8/10, Magnolia CMS, JCR (Jackrabbit), microservices, domain-driven design, RESTful HTTP, HATEOAS, Spring Boot 2, Guice, Groovy, Spock Framework, Handlebars/Freemarker, Maven, Gradle, Git (Bitbucket), Docker Swarm, Docker Machine, Kubernetes, Google Cloud Platform, Bamboo, Jenkins, nginx, Let’s Encrypt (certbot), Stackdriver, Load Impact, Commercetools (SDK, IMPEx, GraphQL)