Coding experiments, becoming an engineer
Shipped interfaces, a knowledge app on real data, and the engineering craft between them — one token layer, one thread of ownership.
Client
Personal / Studio
Services
Tech
Tracks
Product-grade UI and a production RAG pipeline — parallel problems that taught the same full-stack muscle.
Shipped
A public portfolio surface plus a working assistant over messy PDFs — both fed from the same motion and type discipline.
Iteration
Each experiment tightened the loop: design tokens, React surfaces, middleware, and how they meet in production.
Chapter 01
.ARK Studio
.ARK is a digital design studio built like a system: precise, modular, and cinematic. The name comes from the .ark file format—a container that bundles structure, assets, and logic into one coherent unit.
.ARK System
Digital Design Language v2.0
Aa
Geist Sans
Display / Body
Ag
Geist Mono
Code / UI
The Agency Interface
The studio needed a face that reflected its internal logic. I built a live agency site that serves as both a portfolio and a capability statement. It uses the same underlying design tokens as the rest of the ecosystem, ensuring visual consistency across all touchpoints.
Hausarztpraxis RAG System
The Logic Engine
Chapter 02
The Challenge
Transforming a “clusterfuck” of physical PDF documentation into a structured, interactive digital assistant. The goal was to modernize the practice’s knowledge base without disrupting their existing workflows.
AI Implementation & Security
The core of this system is a secure bridge between unstructured data and a conversational interface. We utilize AI not just for text generation, but for structural parsing.
Vector Data Parsing: PDFs are ingested and chunked into vector embeddings, allowing for semantic search rather than just keyword matching.
Secure Backend Bridge: A Node.js middleware layer sits between the client and the LLM provider, managing API keys and enforcing rate limits. This ensures that sensitive medical data structure is handled compliantly.
Persona Engineering: The “medical persona” isn’t a prompt trick; it’s a fine-tuned system instruction set that forces the model to cite its sources from the uploaded PDF knowledge base.
> DATA_TRANSFORMATION_PIPELINE v1.0
Outcome
A functional frontend where patients or staff can interact with a knowledge base that was previously trapped in paper form.
View Live SystemThe Results
One coherent build habit across a shipped interface layer and a RAG-backed medical knowledge app — same tokens, same ownership.
- What shipped
A live portfolio surface and a working RAG medical assistant on top of a vector DB — both using the same tokens, type scale, and motion language.
- What improved
Turned 'AI-assisted' from a buzzword into a measurable workflow: structured prompts, repeatable components and a data layer that stays owned by the practice, not the model vendor.
- What I learned
End-to-end ownership is the real differentiator. When design, systems thinking and engineering share one head, the pipeline stops leaking detail between roles.
Next Project