AI adoption for a set time
Interim manager for AI adoption in mid-sized companies.
In brief
You want to introduce AI and need someone accountable for it. Steffen Herr built his own AI platform, herrlichAI, and leads the adoption for a set time: first the process with measurable benefit, then the tool. At the end your team works with it on its own, with clear rules.
AI is supposed to come, but nobody knows where it pays off first. Steffen Herr leads AI adoption for a set time: first the process with real benefit, then the tool. At the end your team works with it on its own, and the rules are in place.
When interim AI adoption makes sense
The pilot changed nothing
A few licences, a workshop, a test. Day to day everyone still works as before, because no process was changed.
Management wants to see results
Competitors advertise with AI, customers ask about it. Someone has to turn many ideas into two or three projects with measurable benefit and deliver them.
Data is scattered
Knowledge sits in emails, folders and people's heads. Before AI helps, it must be clear which data it may use and where that data lives.
Rules are missing
Since February 2025 the EU AI Act requires that staff using AI are sufficiently trained. Who may do what with which tool? Someone has to define and enforce that.
How interim AI adoption works
- 01 · Weeks 1–3
Find processes with benefit
Talk to the departments, record time sinks and error sources. The result is a few projects with a clear metric, such as handling time per quote.
- 02 · Months 1–3
Deliver
Select the tool, connect the data, change the process. The team works on it from the start so it carries the result itself.
- 03 · From month 3
Rules and handover
A policy for AI use, training for the team and an accountable person in-house. After that the AI keeps running without the interim manager.
What this is based on
- Built herrlichAI: an own AI platform for consulting with an AI-supported content pipeline, CRM integration and automated workflows in production
- Built the AI assistant on complexredugyzer.de, which answers questions about the profile and assesses job profiles
- Digital Transformation & AI Expert (IHK) and AI certificates from Anthropic, Microsoft and OpenAI, among others
- Interim CIO at Telelift GmbH in 2025: rollout of Microsoft Dynamics 365 including migration of existing systems
- 13 years as board member and managing director of a consultancy for mid-sized companies
Frequently asked questions
How does an interim manager for AI differ from an AI consultancy?
The interim manager is accountable for the adoption: decides on projects and tools, runs the project and makes sure the team works with it afterwards. A consultancy delivers recommendations, and the company has to implement them.
Which AI project should we start with?
With a process that occurs often, costs a lot of time and can be measured, such as quotes, enquiries or documentation. A visible success within a few weeks carries more weight than a large programme.
Do we need our own developers?
Usually not at the start. Many projects run on ready-made tools with cleanly connected data. Own development pays off once a process has proven itself.
What does the EU AI Act require of us?
Since February 2025 companies must ensure that staff using AI have sufficient AI literacy. On top come obligations depending on the risk of the use case. An internal policy and training are therefore part of the adoption.
A mandate to fill? An answer within 24 hours.
Availability, terms and fit settled up front. Directly, without detours.
Last updated: 11/11/2026