How do we establish a 'Center of Excellence' (CoE) for RPA governance?

All answers to this question.

For a company dealing with "Shadow RPA," the "Federated" model is usually the best transition. A centralized CoE provides the "Core Infrastructure" (Orchestrator, licenses, security guidelines, and reusable code libraries). The individual departments (Spokes) still have their own developers who understand their specific business logic. This creates "Standardization" without killing the "Speed" that made the departments start their own projects in the first place. You must implement a "Gatekeeping" process where every bot must pass a security and code-quality audit before being moved into the production environment.

   Answered 2025-05-20 by Mary Garcia

In a Federated model, who is responsible for "Maintenance" when a shared application update breaks bots across multiple departments simultaneously?

   Answered 2025-05-22 by Paul Walker

• That's the tricky part, Paul. Usually, the CoE handles the "Platform Maintenance" (upgrading the RPA software itself), but the individual "Spokes" are responsible for their specific bot's logic. To handle the "Shared App" problem, the CoE should maintain a "Dependency Map." When a core application like SAP is scheduled for an update, the CoE alerts all relevant departments so they can test their bots in a sandbox. This proactive coordination is exactly why a CoE is needed—it prevents a single IT change from taking down 50 different departmental automations at once.

     Commented 2025-05-24 by Richard Hall

Don't forget a "Bot Retirement" policy. We had bots running for processes that didn't even exist anymore, wasting expensive licenses for months!

   Answered 2025-05-26 by Karen Young

• I agree with Karen. Lifecycle management is key. Jessica, the Federated model Mary suggested will help you rein in those departments while keeping them happy.

     Commented 2025-05-28 by Mary Garcia