The importance of implementing a Centre of Excellence to reduce the operational risks and inefficiencies in uncontrolled structure

Process automation has drastically changed the way companies work. A current trend in the world is the automation of processes, by robots, known as RPA (Robotic Process Automation) which consists of scripts that emulate human activities that are carried out in businesses daily.

This practice has great market adherence, due to the benefits provided by the gain in operational efficiency. Considering that the setup of robots is an agile process, enabling a reduction of personnel costs. In addition, it can still generate quality gains, as compared to humans, it is less susceptible to errors in repetitive processes.

Would you like to have a deeper knowledge of RPA?

However, because it is a relatively new approach for organisations, the use of the RPA initially occurs without well-defined processes, that ensures the mitigation of errors and risks during the construction and operation of the automation process.


The Importance of Governance

The absence of a control structure makes the automation process more agile and less bureaucratic. On the other hand, the non-existence of well-structured governance in the expansion and use of this automation can cause great operational risks.

Because robots are highly flexible and applicable in various contexts, many organisations encourage the development of new automations in all business departments. For this reason, there is a tendency to think that the RPA does not need significant involvement of the Information Technology (IT) team and this can be a big trap.

Robots can access corporate systems and a lack of governance can cause problems that directly affect the company’s operation, such as a critical application crash. As a real example, a robot captured information in an internal portal from a large company in the banking sector. This was shared by some departments, aiming to gain collective efficiency. However, simultaneous use of automation caused several transactions to run in a short time, overloading the portal and subsequently disrupting general access to the system.

To avoid such risks, automation programs require centralised control and robust governance. Without this, they can become a series of disorderly projects; directing robots to suboptimal tasks with overlapping solutions and using different standards of development and technology, which makes scalability and control difficult. In addition, it carries information security risks and can flood corporate systems with junk mail, as the robot performs batch tasks. If there is a problem, it can only be perceived at the end of the process.

Many different mechanisms can be put in place to make sure the robot raises the proper flags if the execution gets compromised. The issue is in the fact, that the robot is not able to do any rational deviations. Meaning that it can get stuck being unable to finish a given task.


Deploying governance through a Centre of Excellence (CoE)

One way to implement the centralised management of automation is through a centre of excellence (as known as CoE).  This department should ensure that best practices are implemented with the development of reusable solutions and that the use of robots does not generate any risk to the organisation. The CoE organisation can be divided into 3 types:

The responsibilities of the CoE vary according to the established governance model, but some of the most common activities are listed below:

  • Leveraging the topic in the company, training and raising teams’ awareness;
  • Evaluating criticality and quality of solutions developed, ensuring adherence to good practices;
  • Monitoring the operation of robots and ensuring SLAs execution;
  • Monitoring and reporting the main stages of the development process;
  • Ensuring the implementation of the governance established processes;

Ensuring continuous processes and tools improvement.


Other important aspects to take notice

1) Stimulating monitoring

Robust monitoring and security policy are critical to ensure that none of the tools and infrastructure involved in the development of robots is harmed and that they do not infringe the company’s policies. It is important to emphasize that, in order for this quality and criticality assessment do not become problematic and discourage its use, it is necessary to scale it well and share its benefits with the company.

It is essential that all automation that has an interface with business systems, subject to crashes due to the volume of accesses, undergo a validation process in relation to the schedules and the expected volume of its execution.

2) Sharing good practices

In order to ensure improved knowledge management and propagation of best practices, the CoE should promote content-sharing initiatives. One of the most effective ways is through code libraries, which boost new developments, reducing rework. In this process, the CoE should not only act as quality assurance of codes. It also has the responsibility to monitor the development of RPA projects, guiding them so that there is more than one automation performing similar tasks. With a structured and monitored flow, the replication of the qualified content increases the efficiency of the business without providing risks to the operation.



The development of RPAs can be carried out in a centralised, decentralised or hybrid way. Regardless of the model chosen, it is of vital importance that there is well-defined governance on the development and operation of robots, in order to ensure their quality, minimise operational risks and avoid rework. The challenge is to balance appropriate governance with process agility. However, it is possible to reach the ideal system by scaling the responsible department (IT or other defined) by creating policies to support development in the business departments.


About the authors

Vinicius Rodrigues is a consultant at Visagio, a specialist in design engineering projects, analytics, budget management, supply chain and operations, in the retail, consumer goods, health, mining and banking sectors.

Djalma Cunha is a consultant at Visagio, a specialist in logistics, supply chain and operations, design reengineering and data analysis. He has worked in the retail, banking and mining sectors.

Monique Reznik is a consultant with Visagio, with experience in design optimisation projects, CSC, budget, optimisation, logistics and analytics in the retail, banking and real estate sectors.