What it Takes to Set Up a CoE
In my last post I talked about why to have a CoE and where to have it. Today, I want to talk about what it takes to set up the CoE. Just to recap, a business should set up the CoE with delivery capabilities and have it execute on real projects that deliver true value to the business. A successful RPA CoE needs to establish the following three verticals:
- Infrastructure, and
- Operations Management
*Governance and Business Continuity are Disciplines That Run Horizontally Across These Verticals
Dev Ops: Most organizations have defined standards for tool sets and methodologies for various Dev Op functions such as collaboration, source code control, project and change management. Sometimes it is a matter of identifying the correct platform for the function, and some-times the requirements call for implementing new platforms to manage RPA projects. Although RPA platforms do come with some out-of-the-box version and access control functionality, generally, it is not sufficient to support continuous integration and thus requires some work to establish tools and procedures.
Infrastructure and Technical Setup: another area where enterprises struggle. Initial implementation of RPA seems very easy. Setup a web-server with a SQL server back-end with some VDIs and you’re off and running for your pilot. However, scaling this for an enterprise is a whole different ballgame. As the adoption of automation increases, so does the dependency of the organization on bots. This requires implementation of a well thought out, scalable, highly available environment that is segregated to address enterprise security, regulatory or compliance mandated data and access segregation requirements. This also requires buy in from other parts of the technology organization in order to establish standard builds and guidelines for Bot ID provisioning. Data retention and audit requirements must be addressed to ensure logs are kept and are reproducible upon request of the auditor
Operations Management: the third vertical where we see organizations requiring help. Most CoEs are run within the technology organization, which initially thinks of establishing an operations center akin to the a traditional production support organization; the Digital Workforce demands a somewhat different model. This model requires different skill sets for addressing technical system and infrastructure issues vs. business exceptions. Business Operation teams are closely involved in the operations center. Ideally, the digital workers and humans should report up to the same operations manager. This means scheduling and business exception management should be done by operational staff. Thinking through this model and establishing a delineation between IT and business responsibilities becomes critical. Our friend John Slagboom has written a whole thesis paper on this.
*Governance runs across Dev Ops, Infrastructure and the Operations center. You need to establish procedures for how you will manage automation demand across the enterprise, provide a central design and best practice oversight and establish change control procedures.
Lastly, *Business Continuity and Disaster Recovery requires close coordination with infrastructure teams as well as routine drills. Standard procedures need to be defined with specific roles and responsibilities assigned.
All of this can be daunting; but our customers have done their research. They know that this is an involved process and look to providers like us to guide them through the implementation. We are pragmatic with our approach. We recommend our clients create a road-map that gets them to a mature state through gradually evolving stages while continuing to allow them to deliver automation to business units as they establish a fully functional enterprise wide Digital Workforce Operating Model.