Software Engineering · Software Architecture · Enterprise AI
I help companies assess AI realistically and use it where it adds value – from business case and requirements through architecture and software engineering to production operations. When needed, I also take responsibility for implementation – independently or together with existing teams.
25 years of professional software engineering. In recent years, my focus has been on AI systems, agentic architectures, and the dependable integration of modern models.
Analysis and practice
Technical capabilities, product marketing, and actual risks are often mixed together in AI discussions. I write about models, architectures, data protection, costs, and lessons from practical implementation.
Why intensive users can receive remarkable value from an AI subscription—and why exceptionally cheap offers still deserve closer scrutiny.
Read articleWhy a European server location is not enough—and why the provider’s jurisdiction, the server location, and the route to the model must be assessed together.
Read articleWhat distinguishes these three model categories, which hosting options they provide, and what enterprises should consider when evaluating control and data security.
Read articleAssessment and implementation
Good AI systems begin with a realistic assessment of the problem. Model selection, architecture, and implementation follow from there.
Not every interesting use case is economically or technically viable. I help assess value, data, risks, costs, and the conditions required for dependable operation.
Model selection, hosting, data protection, and integration cannot be considered independently. I design architectures that fit the use case and the existing system landscape.
I turn ideas, requirements, and business cases into maintainable software systems, including backends, interfaces, AI components, user interfaces, tests, deployment, and operations.
Many AI systems are unnecessarily slow, expensive, or difficult to reproduce. A clearer separation between deterministic processes and tasks that genuinely need a model often helps.
Technical practice
This selection shows how I combine architecture, established software engineering, and modern AI components into production-grade systems.
Development of a .NET-based orchestration platform for AI agents, workflows, guardrails, and integrations with modern agent protocols.
Token-efficient retrieval layer for large codebases with AST indexing, BM25 search, and fast symbol lookup.
Web search and extraction toolkit for reliable grounding in LLM applications.
How I work
Modern software engineering connects business requirements, technical decisions, and operations whose quality remains verifiable.
Goal and business case
Requirements and risks
Architecture and model selection
Engineering and integration
Production operations
Profile
Twenty-five years of professional software engineering form the foundation of my work. Today, I combine that experience with hands-on AI expertise, software architecture, and the ability to turn requirements into production-grade systems.
Profile and experienceCollaboration
I support assessment, requirements, architecture, implementation, and optimization – as a clearly scoped project, technical consulting engagement, or long-term collaboration.