I design and modernise resilient enterprise platforms — from architectural decisions to production-ready implementation in regulated environments. Including AI agent architectures with enterprise governance.
→Available for architecture reviews, platform design, AI architecture and technical leadership (Remote / DACH / EU).
Project experience with
Consulting
Clearly defined analysis and modernisation packages for complex enterprise systems.
AI Architecture
Structured assessment and design of AI agent architectures for production-ready enterprise systems.
Focus
Outcome
Architecture & Modernisation
Independent analysis of existing system architectures — combined with a structured modernisation strategy and robust target architecture.
Focus
Outcome
Security & Identity
Structural analysis of authentication and multi-tenant architectures in production systems.
Focus
Deliverables
How we start
A structured introductory call (30 minutes). You describe your challenge — I provide an initial assessment and outline possible next steps.
Decisions made upfront save months of remediation. Every engagement starts with a clear, documented architecture baseline.
Security is not an afterthought. Zero Trust, OIDC, HSM integration and regulatory compliance are built in from day one.
Kubernetes-native, observable, and deployable in regulated cloud environments including on-premise and hybrid setups.
Services
Focused on the intersection of distributed systems, security, and operational excellence.
Architecture and implementation of production-ready AI agent systems — from platform selection and governance to scaling.
Automation ↑ · Time-to-value ↓
Domain-driven design, service decomposition, API contracts, and inter-service communication patterns that scale without collapsing under complexity.
Scalability ↑, coupling ↓
Keycloak, OAuth 2.0, OIDC, JWT hardening, HSM-backed digital signatures, eIDAS / e-İmza compliance. Multi-tenant identity done right.
Attack surface ↓, compliance ↑
Cluster design, Helm, GitOps, autoscaling — for public cloud and on-premise. OpenTelemetry instrumentation, distributed tracing, Grafana dashboards — because you can't fix what you can't see.
Cloud costs ↓ · MTTR ↓ · Deployment frequency ↑
Event sourcing, CQRS, Kafka topology design, schema registry, consumer group strategies — for high-throughput, reliable data flows.
Data loss ↓ · Latency ↓ · System decoupling ↑
Schema-per-tenant, row-level security, tenant lifecycle management, and isolation guarantees that satisfy enterprise procurement.
Onboarding time ↓, tenant security ↑
Stack
Production-proven. No hype, no experiments in critical paths.
Work
All projects delivered in productive enterprise operation — with a focus on stability, compliance and long-term maintainability.
Challenge
Enterprises face the challenge of deploying AI agents without governance, cost control or compliance capability. Every team builds its own solutions — without shared infrastructure, no auditability and with vendor lock-in.
Approach
As chief architect I designed and built OMEX AI from the ground up: a multi-tenant AI agent platform with a fail-closed policy engine, knowledge fabric (RAG), multi-provider support (OpenAI, Anthropic, Ollama), FinOps module and admin console (95+ pages). Technology: Java 21, Spring Boot, Apache Kafka, Temporal, Keycloak, PostgreSQL 16.
Result
5,043+ backend tests (100% passing), 117 architecture slices, 777 frontend tests. Production-ready enterprise platform with full lifecycle management for AI agents.
Challenge
Deutsche Bank operated its global network management on a legacy QIP infrastructure. Migration to Infoblox was required — with zero downtime, across multiple global data centres, while keeping operations running.
Approach
As Senior Software Developer, I led the core transformation: data migration, API integration between legacy and target systems, and building a central monitoring portal for global network management.
Result
Global migration in live operation with zero downtime. Centralised network management across multiple data centres worldwide — maintainable, scalable and audit-ready.
Challenge
The Federal Office required a modern, legally compliant solution for automated processing of XÖV/xJustiz messages at high volume — with strict compliance requirements.
Approach
As lead developer and technical director, I designed and implemented the EGVP Spring Boot application: automated message processing, secure transmission protocols, and microservices architecture on Java/Spring.
Result
Solo delivery in 4 months — external teams had declined the project as too complex. Full XÖV compliance, complete audit trail, stable under high load.
Challenge
Bethmanbank required a new generation of digital financial products — process-oriented, multi-tenant, and integrable into existing core banking systems.
Approach
As Senior Chief Architect and Full-Stack Lead, I was responsible for concept, architecture, and development of the platform — from data layer to UI, in close coordination with business units and compliance.
Result
Multiple financial products (risk analysis, portfolio review, investment proposal) in a single rollout. Modular architecture — new regulatory requirements implementable without rewrite.
Challenge
KfW Bank required modern J2EE-based intranet applications for loan processing — with high requirements for data security, traceability, and integration into host systems.
Approach
As Senior Architect, I took responsibility for specification, design, and development of several loan processing modules — from requirements analysis through to production-ready implementation on J2EE.
Result
J2EE platform for development loan processing in live operation. Transaction-safe persistence, complete audit trails for all credit decisions.
Challenge
The Federal Employment Agency required a modern platform for secure data exchange with state authorities — event-driven, containerised, and compliant with strict regulatory requirements.
Approach
As Senior Backend Developer, I built microservices with Quarkus and Java 11, integrated Apache Kafka for asynchronous communication, and deployed the entire platform in Kubernetes — within an agile, multidisciplinary team.
Result
OSCI interface between the Federal Employment Agency and state authorities — multiple federal states connected. Kubernetes deployment, Kafka-based processing under high load.
Challenge
The NGN-MC project to modernise telecommunications processes ran into serious difficulties — structural weaknesses and a lack of technical leadership threatened the project's success.
Approach
Starting as Senior Developer, I was appointed Task Force Lead based on proven performance. I stabilised the project, reorganised the team, and drove it to successful delivery.
Result
Project stabilised and successfully delivered. Technical leadership in a critical phase — transition from developer to lead role under time pressure.
Challenge
No integrated platform existed for workflow management, collaboration, and file management in enterprise environments — existing solutions were too rigid, too expensive, or not integrable.
Approach
As founder and lead developer, I designed and built omex (Object Management Explorer) from the ground up: a complete workflow engine, calendar, email, and file management — all in one seamless cloud platform.
Result
Self-built cloud platform that unites architectural thinking, product development and entrepreneurial initiative.
Whether you need an architecture review, a new platform design, or hands-on engineering leadership — let's find out if we're a good fit.
Get in Touch