Routine
AI as an extra window
The tool runs alongside the work instead of inside it. Whoever wants to use it has to rebuild their own workflow every single time. Most stop trying by week three.
Consulting · AI Integration
You don't buy AI. You integrate it.
Between a purchased license and a changed way of working lies real work: on routines, on rules, on roles. That work is our consulting field. We build AI in where your teams work every day: into communication, collaboration and processes. Tool-neutral, with co-determination and data protection from the start.
Tool-neutral
Routines, not rollouts
Documented measurement
The pattern
The rollout went well. Little has changed.
We know this pattern from 20 years of introducing digital ways of working, from SharePoint to Teams to Copilot: the tool arrives, the way of working stays the same. With AI, the pattern is repeating itself, only faster.
Reach
Islands instead of an organization
One team automates impressively, three watch. The knowledge hangs on individuals and leaves with them: out of the project, the department, the company.
Rules
The policy lives on the intranet
A policy exists, but nobody works by it. It was written next to the teams instead of with them, and it doesn't answer the questions that come up in daily work.
What we actually do
Four fields of work, one goal: routines that last
We work close to teams and processes, not at slide level. Every stage has a clear cut: one routine, one team, one verifiable result.
Rebuilding work routines
Team by teamWe take a team's recurring workflows apart and rebuild them with AI: meeting follow-up, documentation, research, service processes. Piloted in real operations, not in a lab.
Putting the policy into effect
From paper to practiceThe policy draft becomes lived practice: agreed with works council and data protection, explained in the teams, anchored in the workflows. A policy only works once it answers everyday questions.
Enabling where the work happens
MultipliersInstead of sending everyone to a training room, we build capability where it is needed: with multipliers, office hours and casework. For enablement at scale, there is our workshop AI at Work.
Tracking the impact
ProofEvery rebuilt routine gets a documented measurement. If an AI Impact Check exists, its assessment serves as the reference point. So you see what the integration has changed, and what it hasn't.
Fields of application
Where AI integration starts in daily work
Communication
Meetings & team communication
Preparation, minutes, follow-up: the area with the most recurring work and often the fastest effect.
Knowledge
Documentation & knowledge work
Reports, summaries, knowledge storage: AI-supported routines that cut search times and even out quality.
Processes
Service & administrative processes
Recurring cases with clear rules: this is where it is decided whether AI speeds up individual cases or carries entire workflows.
Framework
Governance & policies
Usage rules, responsibilities, co-determination: the framework that turns toleration into a solid foundation.
We work tool-neutrally. Many of our clients live in a Microsoft 365 environment, our core business for 20 years: for the tooling side, we work closely with our Modern Work consulting.
The way there
Integration is best started with a measurement
You can start integration with us directly; the first call then clarifies the starting position. The more solid start is the AI Impact Check: it shows beforehand which routines are worth rebuilding.
Stage 1: AI Impact Check
Diagnosis · 2 to 4 weeksAn evidenced picture across six dimensions: what AI achieves today, what blocks it, what it can still do. With prioritized action areas as the starting point for integration.
Go to the AI Impact CheckStage 2: AI at Work
EnableWhen the people come first: the performance workshop enables teams on their own use cases and sets shared guardrails.
Go to AI at WorkStage 3: AI Integration
Integrate · You are hereThe prioritized action areas are implemented: routines rebuilt, the policy put into effect, the change tracked with documentation.
Not every organization needs all three stages. Stages 1 and 3 can also be bundled into one engagement with a clear scope of services, for organizations that want diagnosis and implementation from a single source.
Why prosma
Integration is our craft. AI is its newest case.
Experience
250+ organizations
We have been introducing digital ways of working for 20 years, among others across the German cooperative banking group. The gap between purchased tool and changed work has been our core business since Teams and SharePoint arrived.
People first
Co-determination built in
Works council and data protection are part of our approach, not an obstacle to it. That speeds up projects, because the most common blocker sits at the table from day one.
Our own practice
We work AI-integrated ourselves
What we recommend, we have tried on ourselves first. Our AI policy is public and shows the rules we work by.
Common questions
Before you call
Do we need the AI Impact Check first?
No, but it pays off: the check provides the factual basis for which routines are worth rebuilding, and its assessment later serves as a reference point. Without the check, we clarify the starting position in a first call and begin with one clearly cut pilot routine.
Are you tied to Microsoft?
No. We work tool-neutrally and recommend what fits your environment. Knowing the Microsoft 365 world from 20 years of project work helps, but it commits us to nothing.
Does this replace our IT service provider or internal IT?
No. Infrastructure, licenses and operations stay where they are. We work on the level above: on ways of working, rules and adoption. In many projects, we work together with the internal IT or the service provider.
How long does such an engagement take?
That depends on the number of routines and teams. We work in clearly cut stages, each with a verifiable result; the scope follows from the check's decision paper or the first call. You can stop after any stage.