Getting AI out of the pilot and into Monday morning.

Getting AI platforms into Monday morning. Most AI programmes don't fail at the model or the platform. They fail in the distance between a tool being available and a team working differently next Monday. I lead AI assets and adoption for IBM in the DACH region, and this talk is what that work actually looks like: the unanswered question that quietly kills usage, the ambassadors who carry a platform further than any roll-out plan, and a show format where teams try AI live and admit on camera what failed.

• Why the bottleneck is rarely the technology
• The response-time metric that predicts whether a platform gets used
• Formats that turn early adopters into multipliers — ambassadors, live shows, first reps

KEYNOTE · PANEL · WORKSHOP · EN/DE
FOR LEADERS RESPONSIBLE FOR AI ROLL-OUTS, IT & HR — YOU LEAVE WITH AN ADOPTION PLAYBOOK THAT DOESN'T START WITH A TRAINING CALENDAR

Building your own AI assets — and what it changes about work. I automate my own busywork with agents: a pipeline that turns a voice note into a publish-ready post, one that turns raw footage into a branded video, a brand kit generator, a morning briefing. None of it is a startup; all of it is in daily use. The talk is a live look at how an AI asset gets from a 2 a.m. annoyance to something a whole team relies on — and at the rules I learned the hard way, like letting agents draft but never send.

• How to spot the task worth turning into an agent
• Draft, don't send: guardrails for agents with outside impact
• From personal tool to team asset — what has to change

KEYNOTE · LIVE DEMO · WORKSHOP · EN/DE
FOR ENTERPRISE TEAMS, DEVELOPER CONFERENCES & INNOVATION UNITS — YOU LEAVE WITH A CHECKLIST FOR YOUR FIRST TEAM AGENT

What digital sovereignty really takes in regulated industries. I led the technical side of taking a sovereign AI platform to production for Europe's regulated industries — public sector and financial services — on two cloud providers in the time usually planned for one. The lesson: sovereignty is mostly a legal, contractual and political question, and only partly an IT-security one. This talk separates the checkboxes from the decisions that actually matter.

• Four questions that decide whether an AI setup is sovereign
• Why speed is culture before it's tooling
• What public-sector teams can adopt today, not after the next framework

KEYNOTE · PANEL · EN/DE
FOR PUBLIC SECTOR, FINANCIAL SERVICES & COMPLIANCE-HEAVY ORGANISATIONS

AI enablement that survives the workshop high. Most organisations don't have an AI knowledge problem — they have a first-rep problem. What actually happens in a vibecoding session, what changes about engineering and non-engineering work when anyone can build, and how to keep momentum after the workshop glow wears off.

• A session format that ends with the team's first AI-built artifact
• What stays human: judgment, taste, safety
• Keeping momentum once the workshop is over

KEYNOTE · HANDS-ON WORKSHOP · EN/DE
FOR ENTERPRISE TEAMS STARTING WITH AI — YOU LEAVE WITH YOUR TEAM'S FIRST SHIPPED ARTIFACT