Virtual Identity (VI) is a digital- and technology-driven agency with deep expertise in UX, digital platforms, and AI strategy. As the demand for AI solutions grew, VI began supporting clients not only in large, business-model-shaping AI initiatives, but also in identifying everyday AI opportunities across entire organizations.

To do this effectively, VI needed a structured approach: a way to understand how teams actually work, which tasks they perform, and where AI agents could support them in a meaningful way.

This need became the starting point for VI’s partnership with teamdecoder.

VI recognized a recurring problem in AI implementation: Many organizations want AI agents — but few have a clear understanding of what their teams actually do.

Clients would often ask: “Can you build us an AI agent to make us more efficient?”

Yet without clarity on roles, responsibilities, and day-to-day tasks, even the best AI concepts remained disconnected from real workflows.

VI realized that to create truly useful AI agents, they first needed a human-centered understanding of work. teamdecoder became the method to deliver exactly that.

To establish Team Analysis as a core component of VI’s AI strategy projects — providing a repeatable, scalable way to:

• map real team work

• understand roles and responsibilities

• identify AI-agent potential based on real tasks

• design Hybrid Human–AI collaboration models
  
  

In short:
From “AI ideas” to “AI that actually helps people.”

VI started by testing teamdecoder on their own HR and Sales teams before applying it to client projects.

The process looked like this:

1. Interviews and Team Sessions

Depending on team size, VI conducted:

• 1:1 interviews

• team workshops

to understand:

• what work needs to be done

• team goals

• how tasks are distributed

• how decisions are made

• where bottlenecks occur

2. Making All Roles and Tasks Visible

Teams mapped their full workload — often for the first time — directly in teamdecoder. For early-stage analyses, tasks were also collected in Miro and evaluated manually using dot voting.

3. Identifying AI Potential

Before teamdecoder had a built-in AI engine, VI and the teams evaluated tasks manually:

• Which tasks follow clear rules?

• Which tasks are repetitive?

• Which tasks require context or empathy?

• Where does the team feel the heaviest load?

This human-first process produced the blueprint for what later became teamdecoder’s Hybrid Team Planner.

4. Co-building the Hybrid Team Planner

The collaboration with a large pump manufacturer was particularly important: Their team analysis data helped define how teamdecoder’s AI should evaluate tasks and suggest AI-agent opportunities.

The result: A feature built from real teams, real work, and real constraints.

1. AI agents only work when human work is fully understood.

Teams learned that effective AI requires precise knowledge of:

• what people do

• why they do it

• how work is sequenced

• what the human judgment points are

Without this, AI agents fail to support real workflows.

2. Acceptance of AI increases with clarity.

When employees see their real tasks reflected in teamdecoder, they trust the AI recommendations more — and are more willing to adopt AI agents.

3. The team, not the AI, makes the final call.

Even when teamdecoder rates a task as “highly automatable,” teams may choose to keep it themselves —
because it is:

• core to their identity

• emotionally important

• tied to context the AI cannot yet understand

This insight strengthened VI’s belief that Hybrid Teams = AI suggestions + human judgment.

4. AI agents require a learning phase.

One major realization:
AI agents do not deliver efficiency gains on day one.

They must be:

• tuned

• corrected

• re-prompted

• iteratively improved

Just like new human team members.

This changed how VI prepares clients and scopes AI projects.

5. Teams welcomed AI as relief, not threat.

While works councils often feared job loss, overloaded teams responded very differently: They were grateful. Because the workload is high and constantly increasing, AI was seen as a realistic form of relief rather than a replacement.

Through teamdecoder, VI gained the ability to:

• offer Team Analysis as a repeatable service

• embed human work understanding into AI strategies

• design AI agents based on real tasks

• introduce Hybrid Human–AI collaboration in a structured, transparent way

• build better alignment between teams and AI solutions
  
  

Team Analysis is now a standard component of VI’s AI strategy, especially for broad, employee-wide use cases.

The next step: Rolling out teamdecoder internally so more VI consultants can use it directly in client workshops.