Too many AI initiatives, not enough visibility. We achieved transparency, clear ownership, and faster delivery with a simple approach. A shared Kanban board in Microsoft Teams as the operational backbone, plus a lean management summary for decision-making. The result: fewer duplicates, clearer accountability, and better throughput.

The problem: Many initiatives, no shared picture. If you need to check Excel files, Jira boards, Teams chats to answer “How many AI initiatives are running right now?”, you’re not alone. We saw the same pattern repeatedly: information scattered across tools and teams, unclear ownership, hidden blockers, and no reliable reporting for leadership. In workshops, one question kept surfacing: “Is someone already working on this use case?” Surprisingly often, nobody knew. As the number of ideas, pilots, and implementations grew, the lack of a consolidated view began to slow down execution and obscure progress.

Why AI initiatives need dedicated coordination AI doesn’t fit neatly into traditional project boxes. It cuts across S/4HANA, Ariba, Datasphere, Signavio and beyond. Teams are cross-functional, and the underlying AI portfolio evolves continuously. Without a shared overview, it’s easy to duplicate efforts, bury valuable knowledge in silos, and starve leadership of the signal they need to make prioritization decisions. Coordination becomes the bottleneck. That led us to a simple question: How can we make every AI activity visible—without adding yet another project management tool?

 

Our Solution: A Teams Kanban Board for AI Activity Management

Instead of introducing another complex application, we decided to leverage a tool that everyone was already using every day: Microsoft Teams.

By creating a dedicated AI Activities Kanban Board, we established a single source of truth for all AI-related initiatives—from the first idea through productive deployment.

Operationally, the board changed the rhythm of collaboration. Teams now have a single place to see what’s in flight, what’s blocked, and who is accountable. Stand-ups focus on movement between columns rather than gathering status. Duplicate work shows up early. And because the board is embedded in the tools people already use, the barrier to adoption is low: no new logins, no new training, just a clear and shared way of working.

 

Beyond the Board: Reporting for Management and Stakeholders

The board is a working tool for the team. Management needs something different: condensed, decision-ready information. That is why we built a Management Summary on top of the board — structured into three sections:

 

1. Overview — A status distribution across all phases, the current Top 3 priorities (e.g., Deduction & Dispute Agent, Sales Order Creation Agent, Joule for Consultant), and AI unit consumption against the contracted budget.

2. Efficiency Metrics — Four indicators that answer whether we are getting faster and better: completed tasks, average cycle time, current work in progress, and a 6-month WIP trend. 

3. Outlook — Upcoming milestones and the backlog broken down by category (Joule, Embedded AI, Custom AI, AI Agents) to support prioritization decisions.

 

How It All Connects

The Kanban Board is the living, daily working tool. From it, we export data regularly and generate the Management Summary, either as a PowerPoint deck or through Power BI dashboards for automated reporting. Additional context (e.g., strategic commentary, risk flags) is added manually where needed. This ensures that:

• The project team has a real-time collaboration tool.
• Management receives condensed, decision-ready reports without needing to interpret raw board data.

 

Final Thoughts

AI initiatives move fast — and that pace is only accelerating. The approach described here is deliberately simple: a Kanban Board as the operational backbone, and a lean reporting structure to bridge the gap to management. No heavy tooling, no lengthy rollout. What it does require is discipline — keeping the board alive, reporting consistently, and treating transparency not as a reporting obligation but as a governance principle.

If you are currently managing AI activities in spreadsheets, email threads, or informal conversations, this might be the right moment to take a step back and set up a structure that scales with your ambitions.

I am happy to discuss further or share experiences — feel free to leave a comment or reach out directly.