A 15-year-old, €100M print platform rebuilds itself around AI - while the business keeps running.

- Use case · Helloprint

Helloprint is one of Europe's largest online print platforms. They've used machine learning for thirteen years and got serious about AI in 2022. This is the story of the moment they stopped applying AI to the old organisation - and started redesigning the organisation around it.

Where they actually were.

This is a fifteen-year-old company with real customers, real suppliers, real margins, and real legacy. The kind of operation where 'apply AI' is easy to say and very hard to do without breaking something.

The turning point.

For years Helloprint layered AI on top of the existing structure. It worked in pockets, and then plateaued. The bottleneck turned out to be the operating model around the tools: departments, project managers, three-week cycles, tickets between idea and shipped code. That structure simply couldn't keep up with what the models could now do.

We came in to support Helloprint in their AI Transformation, alongside the leadership team and a growing group of people inside the business who wanted to build.

Phase 1 - what didn't work.

Most case studies skip this part. We won't. A dedicated central AI team ran for about a year, staffed with strong people but disconnected from the business. Projects took too long or stalled outright. A flagship content tool was already outdated by the time it shipped six months in.

The lesson that matters for other leadership teams: an AI team without business context produces the wrong thing, slowly. The context has to live inside the people doing the building.

From departments to processes.

The org chart was redrawn around eight core business processes instead of around departments. Process owners replaced department heads. The PM layer largely disappeared. 'Everyone is a builder' became literal, with one nuance: process owners still lead. There is structure - it just sits closer to the work.

What disappeared was the coordination layer between idea and ship: tickets, sprints, three-week cycles. The distance between 'we should do this' and 'this is live' collapsed.

The heart - the SuperIntelligence platform.

An internal platform that wires the AI assistant into every system that matters - data warehouse, CMS, CRM, collaboration, analytics, commerce - through a governed integration layer with role-scoped permissions and a full audit trail. It's the difference between 'we use AI' and 'we have a company brain.'

Skills - what the organisation actually built.

Around eighteen production skills, built by the people who use them. Finance, customer experience, quality, operations, design. Concrete people, doing concrete work.

The cadence.

Every Tuesday 14:00 - an open training hour. One-on-one champion deep-dives. Builders are celebrated publicly. The cadence is the system. 'It takes a village to move people from playing around to genuinely changing how they work.'

What changed in the business.

Production skills shipped across finance, customer experience, quality, sourcing and commercial - built and owned by the people who do the work. Manual ticket handling dropped meaningfully. Time-per-quality-review collapsed. Whole departments shifted from doing repetitive work to instructing agents that do it for them.

Hiring is being rethought role by role, with some openings deliberately tested as agent-first before a human is added. Adoption moves in waves; usage dips and recovers as the model evolves. What's durable is the muscle: the business can now design, ship and govern its own AI work.

What Dissel contributed.

As the AI Growth Program Committee, we supported Helloprint in executing AI within their business. From a use-case strategy mapped on their core processes to building AI products, and learning along the way.

In the second half of the project we contributed to rolling out their AI platform (Helloprint Superintelligence) and supported with the rollout sequence, the org logic, the governance, and the rhythm that keeps it moving. We built the governed integration layer that lets agents act safely on real company data, and the brand-aware tooling that keeps output on-voice.