About – Michael Ikemann

I help companies build AI systems that work in the real world — not just in demos.

I started developing at the age of seven, sold my first software at ten, and by sixteen was an expert in real-time 3D graphics, developing for the Riva TNT and 3dfx card generation. At 11880 I built and technically led Germany’s biggest mapping and routing platform, serving 10 million users — and was one of the pioneers of iPhone development, with a long-time #1 App Store app.

My background is not one industry. It is the repeated translation of expert knowledge into systems that actually run: from commercial computer games over navigation and self-driving car systems to industrial computer vision, and most recently the LLMing platform — an enterprise AI platform I designed and built, used in production around the globe. The pattern is always the same — the expert knowledge is the hard part, and software that captures it makes organizations move faster.

Today I lead a global team of data scientists and AI engineers as Head of Data & AI at Lechler GmbH, and I take on selected consulting engagements through ikemann.ai — with deep expertise in document processing, computer vision, and agentic AI.

Domain First Development

The people who understand a business problem best are rarely the ones who can change software fastest. Engineers, planners, sales and service experts know the edge cases, calculations, supplier rules and operational shortcuts — but in classical organizations they can’t turn that knowledge into maintainable software.

My work closes that distance: I provide the platform, guardrails, runtime and learning foundation — domain experts provide the actual intelligence of the business. The result is often not even “an AI product”: it’s a calculator, a dashboard, a workflow helper or an MCP tool that simply executes what the expert knows. And because the systems verify themselves, every failure makes the next attempt stronger.

Track record

2026 – today

AI Consultant · ikemann.ai

Freelance consulting for document processing, computer vision, and agentic AI — designing and delivering solutions that automate complex workflows, extract intelligence from visual data, and let AI agents act autonomously and reliably.

2023 – today

Head of Data & AI · Lechler GmbH

Leading a global team of data scientists and AI engineers and building practical Data & AI capability around domain experts — including quality assurance solutions for Lechler’s production. Designed and built the LLMing platform, which Lechler uses to bring multi-model AI chat, agentic document workflows, and AI-assisted mail and news intelligence to every employee — with tools such as llming-com and llming-stage empowering employees to build their own AI agents and customized tools.

2018 – 2023

Lead Data Scientist · Zühlke Group

Industrial computer vision: quality assurance, process surveillance, and document processing — for Bosch Cognitive Services, the PDR Team (e.g. automatic hail-damage detection), and Lechler, among others.

2010 – 2018

Head of Software Engineering Mobile · LOGIBALL GmbH

B2B geoinformatics: fleet management and optimization, navigation systems, aerial-image processing, and supporting self-driving car development.

2005 – 2010

Technical Project Lead · 11880 Internet Services AG (klickTel)

Technical lead for Germany’s biggest maps & routing portal — a high-performance mapping and routing platform serving 10 million users, and a long-time #1 App Store app.

2002 – 2005

Freelancer · Ikemann IT Services

Custom software development and network infrastructure across practical customer projects.

2000 – 2002

Senior Software Developer · VX Software GmbH

Commercial computer game development on Sternenfahrer von Catan (Starfarers of Catan) and Schiene & Straße (Trains & Trucks).

1989 – 1998

Roots

Participated in the German demo scene, Usenet and IRC; first sold software in 1992 and participated in the book Computer Games & Development in 1998.

Open source

I believe in building in the open. The LLMing stack, agentic tooling like yikes! and IsoPod, and a range of vision and UI libraries are available on GitHub — see the full project list.

Michael Ikemann