How the world’s largest inland port went from shadow AI to official AI.

Who are you?

I am Felix Musolf, Head of Digitalization & AI in Corporate Development at duisport, the largest inland port in the world.

What does your company do?

As a trimodal logistics hub and one of the most important nodes for trade flows in Europe, we operate across three business areas: logistics, packaging, and infra- and suprastructure. My department, Corporate Development, ensures that we keep pace internally, both technologically and in the skills development of our employees.

Since when have you been a customer?

We have been partners with NENNA for about 1.5 years, starting with a pilot of 25 volunteers, and today it is in use across almost all specialist departments.

1. What "superpower" does NENNA give you?

2. What do you use the platform for in your daily business?

"From HR, IT, and Legal to Procurement and Event Management, NENNA runs across all teams. Pre-analysing contracts, researching market and trend analyses, developing campaigns, preparing price negotiations, supporting code. It varies depending on the department, but it is in daily use everywhere."

3. What is your personal WOW feature?

"Our WOW feature? The privacy button. That is what we call it internally. It allows you to actively control which data flows into which models. That sounds highly technical, but it is actually a psychological feature: legal colleagues, for example, suddenly work with AI without any hesitation."

4. What does NENNA do for your company?

NENNA has brought us from shadow AI to official AI. Previously, colleagues used ChatGPT and the like. We quickly realised that data was leaking. Today, more than 13 departments work openly with AI because they know: it is data-secure, it is permitted, and it works. And all this without us having to set up a GPT ourselves, sort out liability, or clarify operational responsibility. That is all taken care of."

5. Which NENNA model do you have in place?

"Not just one model. Several. Depending on what is needed. For quick, everyday tasks, lightweight models are completely sufficient. For complex analyses in Legal, Data, or BizDev, we rely on more powerful ones. Some we have even deliberately deactivated. Not because they are bad, but because we simply do not need them. That is precisely the point: we only pay for what we actually use. The alternatives we looked at would have been three times as expensive for the same range of functions."

6. What would you miss without NENNA?

"To be honest: peace of mind. We would miss the secure feeling that AI usage is no longer a grey area for us. The alternative would be to return to silence, with people secretly using ChatGPT and the like again. Or we would build something ourselves, with everything that entails: licences, operations, liability. Nobody wants that. Technically, a switch would be feasible. But people would not want to give up that security anymore. It is no longer just a feature. It is the standard."

7. How would you describe the collaboration with NENNA?

Very user-friendly, very direct, highly constructive, solution-oriented, pragmatic, flexible, easy-going.