Senior Engineer for AI Integration, System Architecture, and Production-Grade SaaS

I am a freelance senior engineer for AI integration, system architecture, and production-grade SaaS. For more than 25 years, I have built applications and platforms where business logic, permissions, data models, and operations need to work together reliably.

My focus is not AI for its own sake. I integrate LLM capabilities into real products with clear system boundaries, server-side orchestration, guardrails, cost control, and dependable delivery in B2B and enterprise environments.

25+ years of engineering experience · OpenAI / Azure OpenAI · RAG, guardrails, structured outputs · RBAC, entitlements, audit logs · Remote from Germany for projects across Germany and Europe

Discuss your project See Valutra case study

Also open to other demanding implementation work beyond AI when architecture, technical depth, or end-to-end delivery responsibility matter.

What teams bring me in for

When AI needs to work as a controlled part of a real product rather than as a demo. When permissions, compliance, cost control, and operational stability matter from the start. And when architecture decisions need to support long-term product evolution.

Integrating AI into real systems

Embedding LLM capabilities so they work cleanly with product logic, data models, and operational requirements.

Making architecture enforceable

Designing roles, entitlements, auditability, guardrails, and clear system boundaries from the beginning.

Owning delivery through operations

Treating implementation, quality standards, deployment, and long-term maintainability as one engineering responsibility.

Service Focus

Three areas where I currently create the most leverage for teams and products.

AI Integration for Real Systems

LLM integration, RAG, server-side orchestration, guardrails, structured outputs, and token or cost control for applications that need reliability rather than hype.

System Architecture & Platform Design

Architecture for B2B SaaS and platform systems with clear boundaries, RBAC, entitlements, audit logs, maintainable data models, and strict server-side enforcement.

End-to-End SaaS Delivery

From data model, API, and UI through CI/CD, deployment, quality controls, and operations. The goal is not just delivery, but ownership.

Selected Systems & References

Examples of the kind of responsibility and system context I work in.

AI assistant platform in a testing and certification context

Production AI assistant platform for multiple business units with integration into existing enterprise systems and internal knowledge sources.

Integrated Azure OpenAI, DeepSeek, and enterprise data into a scalable RAG architecture using pgvector, embeddings, and semantic search
Built streaming chat with dynamic context enrichment plus defined context, prompting, and retrieval strategies for reproducible results
Implemented guardrails, validation mechanisms, multi-tenant architecture, and cost control through token analysis and optimized request strategies

Valutra – deterministic product logic as a system foundation

Production-grade financial forecasting SaaS with deterministic, rule-based cash-flow simulation and strict server-side enforcement of roles, entitlements, and limits.

Owned end-to-end: product logic, UI, API, data model, and operations
Not an AI product, but a credible example of hybrid architecture with deterministic core logic and optional AI extension points
WCAG 2.2 AA, auditability, and long-term maintainability as system requirements

View Valutra live → Technical details

Azure platform architecture & self-service infrastructure

Evolved an existing Azure self-service platform into a modular, cross-service platform and previously designed and implemented an application for group and VM management.

Defined generic interfaces, service concepts, and extension mechanisms for adding new infrastructure services across multiple teams
Integrated Microsoft Entra ID, Azure Compute API, and stable backend APIs while delivering core frontend and backend components with React, Next.js, and Node.js
Provided technical sparring, mentoring, CI/CD pipeline work, and container-based deployments via Azure DevOps and Docker

Also open to other demanding implementation work

Beyond AI integration and architecture, I am open to other demanding implementation work where technical depth, product context, and reliable execution matter.

Extending and stabilizing existing B2B systems
Delivering new platform modules, SaaS capabilities, or integrations
Modernization work with clear delivery and operational responsibility

Additional references across e-commerce, consent/tracking, learning platforms, and earlier banking or logistics projects are available on request.

Architecture, AI integration, or another demanding implementation project?

If an existing product needs meaningful AI integration or a new system needs a reliable foundation from day one, let's discuss architecture, scope, and the next practical step.

Discuss your project