01
Applied AI systems
- Local LLM inference and model evaluation
- Persistent AI state and capability boundaries
- Reasoning engines and benchmark/evaluator workflows
- Secure worker / inference integration
Résumé
Computational Engineering Master's student with a Mechanical Engineering background, building software products across applied AI, backend/data, Linux infrastructure and delivery.
Profile
My work spans consumer products, transactional business systems, local AI infrastructure and the engineering systems behind them. I'm strongest when a problem crosses application, data and operations boundaries.
I use ChatGPT and Codex extensively for implementation, research and debugging, while retaining ownership of product intent, architecture, constraints, acceptance criteria and real-system verification.
01
02
03
04
What I can own
My strongest evidence is not one isolated framework. It is taking a bounded product or systems problem from architecture through implementation, verification and delivery.
AI-enabled systems
Inference, durable state, evaluation, capability boundaries, client integration and local/private deployment.
Data-backed products
APIs, PostgreSQL, transactional state, provenance, admin/operator workflows and responsive product surfaces.
Delivery & operations
Linux, Docker, CI/CD, exact releases, observability, infrastructure boundaries and rollback/recovery thinking.
Selected evidence
These are selected systems, not a complete repository list. Each one demonstrates a different part of the engineering range.
Applied AI / developer tools
Diagnostic reasoning platform with a separate Intelligence Lab for controlled benchmark, evaluator and engine evolution.
View case study →Local AI systems
Persistent private assistant architecture with PostgreSQL state, dedicated GPU inference, mTLS transport and bounded capabilities.
View case study →Backend / transactional systems
Farm-to-consumer platform covering traceable supply, pricing, concurrency-safe order commitment, fulfillment and settlement foundations.
View case study →Consumer product / recommendation intelligence
Discovery and community product with canonical catalogue data, provenance, privacy-aware domain modelling and research-backed recommendation evidence.
View case study →Education
The academic path adds systems, modelling and computational context; the project portfolio provides the practical software, AI and infrastructure evidence.
Open to focused freelance projects now and permanent roles in Germany across Technical Product Engineering, Applied AI / AI Systems, Backend / Full-Stack, Platform / DevOps and adjacent technical roles where cross-layer ownership is useful.
Contact / availability