Kaufmännische Praxis · AI-assisted workflows & digital operations
I build workflows around practical tasks: finding a document, checking a handoff, producing editable files, or deciding what a learning system should do next. The projects below include public evidence you can inspect in your browser.
| Project | The practical problem | Open the evidence |
|---|---|---|
| Effective PDF | Find PDF content through Markdown, trace results to source pages, and detect stale indexes or changed documents. | Search the demo archive → |
| MoveNLearn | Turn a structured enquiry into an explicit human-review task; connect research, content and document production. | Explore the intake handoff → |
| Editable Presentation Pipeline | Produce editable PowerPoint files, PDFs and speaker scripts while checking export quality. | Compare designs and speaker scripts → |
| UniPrep | Keep instruction evidence, assessment and the next learning step consistent across sessions. | Follow a complete learning path → |
| FitnesHermes | Validate nutrition and activity data before a personal assistant uses them. | Inspect missing and stale data → |
I define the problem and requirements, design the workflow, direct implementation with AI coding assistants, and review the outputs. Each repository identifies its own implementation, external libraries and the limits of its public demo.
My background combines commercial work in real estate administration with website QA and business-development work during an Erasmus+ internship in Valencia. That practical context shapes the projects: explicit inputs, traceable decisions and outputs another person can check.
All public examples use synthetic data. The case studies distinguish recorded demo results from live integrations; personal study, customer and health records are excluded.
Learning Apps · Markt- und Preistheorie is a small German economics learning prototype with quiz formats, diagrams and browser-local demo results. Open the app →