Experience from the field
From first workshop to production, in five months
A healthcare network with more than 2,000 employees faced a classic problem: leadership wanted to adopt AI, the IT department had concerns about data privacy and compliance, and the business units did not know where to start. Three factions, three languages, no shared foundation.
Leadership workshop, structured agent audit across 4 departments, governance framework developed with IT. First agents built in administration and cross-departmental coordination, built together, then owned by the business units.
First productive AI agents deployed across the organization. A governance framework the IT team stands behind. An internal team capable of developing the next agents on their own.
20 concrete agent opportunities identified
A mid-size MedTech company with 800 employees had already discussed AI internally, but without result. Too many ideas, too little prioritization. Leadership wanted clarity: what is realistic? What creates real leverage? Where do we start?
Structured agent audit across all relevant departments. Each opportunity evaluated by benefit, feasibility, risk, and effort. Result: a prioritized roadmap.
20 concrete agent opportunities: 7 quick wins, 8 medium-term projects, 5 strategic initiatives. Estimated 30–70% productivity improvement in marketing and sales.
AI as a creative tool, not a threat
A large media company with multiple editorial teams faced a typical leadership challenge: employees saw AI as a threat to their jobs, not as a tool. Leadership wanted a culture of experimentation, without forcing the change.
Executive workshops for leadership across all divisions: each participant built their own agents for their specific work area. The focus was not on technology, but on the question: which tasks would you like to delegate? What would you do with the time you gain?
The attitude shifted. Executives who came in as skeptics became internal champions. Concrete pilot projects emerged from the workshops, initiated by the business units themselves, not ordered from above.
AI strategy for deal sourcing and portfolio analysis
A private equity firm with more than 30 portfolio companies across Europe wanted to understand where AI agents create real value, without disrupting established investment management workflows. Deal sourcing, portfolio monitoring, and exit preparation were still largely manual.
Workshop series with investment managers: analysis of deal sourcing, portfolio reporting, and market monitoring. Three areas with the highest automation potential identified.
Three prioritized areas of application with a clear implementation roadmap. A first agent for portfolio monitoring is built and in testing. The goal behind it: take research off the investment managers' desks so more time is left for strategic analysis.
Finding the right tenants out of 500 inquiries per listing
A brokerage for residential and investment properties managing several thousand rental units. For sought-after listings, 500 to 600 prospective-tenant inquiries come in. Per listing. Sorting and prioritizing was done by hand, on gut feeling, with the landlord's criteria living only in the broker's head. The best tenants dropped out because the answer came too late. Hiring more staff was not an option.
It started with a workshop. The whole team built their own first AI applications in a single day, live, with no prior experience. From the ideas, leadership picked the first production project. Since then, a prioritization agent has been taking shape. It matches incoming inquiries against the landlord's criteria and returns a justified ranking, a traffic light from green to red. The agent pre-sorts, the decision stays with the human. The first prototype was built by one of the firm's own employees.
Currently in progress. The prototype is built, the pilot follows. The shared goal: significantly faster pre-sorting per listing, a traceable ranking instead of gut feeling, and a team that develops the agent further on its own.