Robert Glaser
Head of Data and AI, Principal Consultant
As Head of Data & AI at INNOQ, Robert helps boards and development teams run the full AI marathon—from use-case discovery and portfolio scoring to pilots and production—without getting stuck in proof-of-concept-hell. He designs operating models where agent tooling, eval loops, and compute realities reinforce each other so ambitious ideas actually ship. He hosts the podcast “AI und jetzt”, takes his techno-philosophy blog and newsletter on stage in conference talks, and keeps prodding leaders to think of AI as a foundational technology instead of a mere tool. In his spare time he chases single-origin beans and new restaurants, insisting palate calibration is just another form of model evaluation.
Podcast “AI und jetzt” (German)
- MCP: Copy-Paste adé, People Management für KI, Unternehmens-App-Stores
- AI-Burnout, High vs. Low Agency, ein Sekretariat für das mittlere Management
- AI-Überforderung, Konzernträgheit, Deutsche Ingenieursmentalität mit Gästin Barbara Lampl
- Abundance Mindset, Werkzeuge schärfen, KI-Transparenz mit Gast Marcel Weiß
- Currywurst aus Gurken, Text ist Design, Deutsche Bräsigkeit mit Gast André Hennen
Books
- RAG: Retrieval-Augmented Generation – Die Architektur zuverlässiger KI, zusammen mit Co-Autoren Alexander Kniesz, Marco Steinke, Hermann Schmidt
- Datenbanken: Datenmodelle, Skalierung, Cloud, zusammen mit Co-Autor Marco Steinke
Current Technical Talks
Talks can also be held in-house. Feel free to contact me.
- Integrating AI Agents with MCP – Bridging Distributed Systems and Real-World Data
- RAG: The Architecture of Reliable AI, annotated Talk+ for later reference
- No More Excuses: Automating Accessibility with GenAI
Current Keynotes and Impulse Talks
Inspiring, pointed, and full of perspective – ideal for panels, keynotes, and thought-provoking shifts. Custom talks (including in-house sessions) are always possible – happy to discuss ideas anytime.
- High Agency as a Survival Strategy: Navigating AI-Driven Transformation (also available tailored for IT and Development Departments)
- Stop “Evaluating” AI – Start Building Systems that Work
- Bicycle of The Mind – How AI Changed the Way I Learn
- Stillstand im Fortschritt: Der paradoxe Umgang mit KI, als Talk+ auch zum Nachlesen verfügbar
Current articles
- Agentic Pelican on a Bicycle
- The End of Shouting: Prompts as Programs in GPT-5
- Fire Your Fetch Tool — Let your Agent Use a Browser
- Claude Skills, anywhere: making them first-class in Codex CLI
- Die leidigen Use-Case-Schubladen (German)
- CIO.com: Have we reached the end of ‘too expensive’ for enterprise software?
- What tracks do we leave behind with technology? How monsters can teach us about responsibility (Reinforcement Learning)
Newsletter: This Week in Data & AI (German)
- Wenn der Agent das Labor betritt
- Karpathy statt Tatort: Kein Entwarnungssignal
- Die GPU im Keller: Was passiert, wenn High-End lokal wird? (KW 41, 3.–12.10.2025)
- KI-Projekte scheitern auf dem Weg zur Kantine (KW 40, 27.9.–2.10.2025)
- KI fließt und wir brauchen Ventile (KW 39, 20.–26.09.2025)
- Der kleine Freund in der Box hat jetzt einen Arbeitsvertrag (KW 38, 15.–19.9.2025)
- Agentik braucht Struktur, nicht mehr Prompt-Zauber (KW 37, 8.–14.9.2025)