Hamburg / Tech / AI

Hi, I'm Stephan.

Tech Founder & Software Engineer

I build things and write about what I learn.

Interested in Tech and Business, AI systems and how language models actually work.

Explore Work Read Writing Open Blog

What I build

A compact view of the products, tools, and company work that show how I think and ship.

Featured experience

verlingo

Founder and Managing Director · 10+ years

Built and scaled a B2B SaaS company in HR-Tech, combining entrepreneurship, applied machine learning, product strategy, business development, and customer-centric execution to simplify complex HR workflows.

Led product, business development, and customer execution across a decade-long SaaS journey.
Turned complex HR workflows into clearer digital products for business users.
Grew verlingo to 400+ enterprise customers across Germany, AI/ML-first from day one.

AI Media & Storytelling

Kreativbuch

A live product for creating personalized children's books with custom characters, stories, and illustrations, available as PDF, softcover, or hardcover.

AI StorytellingPersonalized BooksImage Generation

Visit

Automation & Business Engineering

Ausschreibungszentrale

A live product that turns scattered public tender portals into clearer dealflow: relevant matches, prioritization, alerts, and AI-assisted application workflows.

Tender AutomationSaaSSQLite

Visit

AI Document Analysis

Zeugnisprofi

An AI-assisted product that helps people understand employment references by translating coded certificate language into clear grades, explanations, and improvement suggestions.

AI AnalysisHR-TechDocument Workflows

Visit

Open Source & Tools

Scanic

An open-source document scanning library for document edge detection and perspective correction in browser and Node.js environments.

TypeScriptWebAssemblyRust

Visit

AI Media & Storytelling

Luniko

An AI media and storytelling project focused on personalized stories and immersive audio experiences.

AI MediaAudio PlaysPrompt Systems

Visit

Writing

Recent writing

Things I've been reading, building, and thinking about.

2026-06-01

How LLMs Got Faster Without New Hardware

The training side of AI gets all the headlines. But in the last year the inference side has been quietly transformed. KV cache compression, rotation-based quantization, and multi-token prediction are why your local model runs twice as fast today as it did twelve months ago.

Read

2026-06-01

The NanoGPT Speedrun: From 45 Minutes to 81 Seconds

A GitHub repo nobody talks about might be the clearest window into how AI training actually got so much faster. Here is the story of modded-nanogpt and the people racing to train a language model as fast as physically possible.

Read

2025-08-21

The Sky Is Patient

I once bet we'd hail an air taxi by 2027. Today, I'm less confident and here's why.

Read

Open for focused conversations

Contact

For focused conversations around software engineering, AI systems, and product work.

Email me