Agentic AI in action: our first Ship-IT spinoff
At ACA, Ship-IT Days are no-nonsense innovation days.
One day, out of the usual routine, to focus entirely on bold ideas, rapid prototyping, and hands-on experimentation. No lengthy processes. No “we’ll get back to it later.” Just building something that works today.
During a typical Ship-IT Day, ACA colleagues form cross-functional teams and dive into a topic or problem. By the end of the day, they present a first version of what could become a new feature, product, or service. It’s not a hackathon. It’s not a conference. It’s a working sprint to test real ideas, fast.
If you’re curious about how things looked last year, you can check out the 2024 edition here.
A focused spinoff: Agentic AI
While the next full-scale Ship-IT Day is set to return in September 2025, we recently tested something new: a focused, smaller-scale version of it. A spinoff focused on the topic: Agentic AI.
This first edition took place at our ACA Group office in Antwerp and was initiated by Stijn Van den Enden (CIO/CTO of the ACA Group) and Alexander Frimout (AI Expert & Unit Manager). Four teams rolled up their sleeves and teamed up with clients to take on real challenges, from logistics to ESG data and even smart reception flows. Hands-on, practical, and concrete. Exactly what a Ship-IT day should be.
Why Agentic AI?
AI is everywhere: in meetings, tools, podcasts, the news and in that one meeting where someone says “we should do something with AI” and then never follows up. But what can it actually do for your organisation, right now? How do you move beyond the talking phase?
That’s what we wanted to figure out during this Ship-IT spinoff, where we zoomed in on Agentic AI: AI that doesn’t just wait around for input, but observes, makes decisions, and takes action (from inbox handling to generating reports or processing documents without human babysitting). Not to simply replace people, but to take some of the brain-numbing tasks off their plate.
That’s where Agentic AI comes in. It goes further than traditional automation. These systems don’t just react, they observe, make decisions, and act on their own:
- an AI that processes incoming emails and sends accurate replies
- one that digests unstructured ESG data and outputs usable structure
- or an AI receptionist that actually opens the barrier and links up your billing system
Not to replace people, but to clear out the repetitive tasks, so your team can focus on more strategic work.
Four clients, four cases, one day
The teams tackled a wide range of challenges:
CXL: AI Reception Agent
- Website: www.cxl.be
- What they do: CXL offers high-end automation and software solutions for hospitality, including hotels and parking systems.
How much can you take off the plate of a front desk? This team created a digital receptionist that can handle bookings, open the parking gate, and even connect to billing systems, all through natural language. The agent was up and running by the end of the day, and took home the unofficial “winner” badge.
___________________________________________________________________________
amfori – AI-powered ESG Data Cleaning
- Website: www.amfori.org
- What they do: amfori is a global business association for open and sustainable trade, helping companies improve ESG performance in their supply chains.
The amfori ESG Risk Compass depends on data from dozens of external sources, all in different formats. Processing that data is time-consuming and technical. The team built a drag-and-drop interface powered by AI, enabling sustainability experts to upload raw data and automatically transform it into structured, ready-to-use output.
___________________________________________________________________________
Katoennatie – Email Response Automation
- Website: www.katoennatie.com
- What they do: Katoennatie is a global logistics and supply chain company operating in port terminals, warehouses and industrial services.
“What’s the status of shipment 12345XYZ?”
Normally, a human would read the mail, search the system, and type a response. This team built an AI agent that does exactly that, identifying reference numbers, pulling live data, and generating accurate replies. A solution that saves time and gets smarter with every use.
___________________________________________________________________________
TVH – Intelligent Data Cleaning
- Website: www.tvh.com
- What they do: TVH is a global parts specialist, providing spare parts and accessories for material handling, industrial and agricultural equipment.
TVH is migrating thousands of legacy data records into a new system, but the data is messy: duplicates, missing info, inconsistencies. This team explored how AI can help clean and enrich the data automatically, spotting patterns and suggesting corrections in real time.
___________________________________________________________________________
The result? Real progress
By the end of the day, every team had a working prototype to show. No promises, no theories, just working code and real feedback. For some clients, it was a first glimpse into what AI could mean in practice. For others, it was the missing push to move from “we should really look into this” to “let’s build this out.”
That’s the power of the Ship-IT format: Clear scope. Small team. Just enough time pressure.
And most importantly, the freedom to build without waiting for the perfect moment.
*Big congrats to our CXL Ship-IT Day team, who were voted the winning project of this edition.
Want to know more about Agentic AI Hackaton?
Got questions about AI? Reach out to Alexander!
