Automation both reduces the cost of traditional IT (fewer resources used) and avoids the errors inherent in manual development. Coupled with the cloud, it enables greater agility. This is the promise of RPA (Robotic Process Automation). As long as you're methodical, and your IT security is reinforced.
Having achieved a certain level of cloud maturity, the majority of companies (84%) are focusing their strategy on multicloud, according to a study by RightScale/Flexerade. Adoption of private clouds is slowly increasing, with spending expected to rise by 8% in 2019 compared with last year. Hybrid cloud usage continues to grow, with enterprise adoption rising from 51% in 2018 to 58% in 2019.
In a nutshell, the cloud has become an essential link in the digital transformation of companies, whatever their size or activity.
Accumulation of multiple layers
As organizations integrate the cloud into their digital strategy, they are beginning to take an interest in the various functionalities linked directly or indirectly to the cloud.
Such is the case with process automation in the cloud. On paper, the principle of RPA (Robotic Process Automation) is simple and attractive. RPA software automates repetitive, rules-based processes usually performed by people sitting in front of computers, to use the definition given by David Schatsky, Managing Director at Deloitte.
This solution represents an opportunity to simplify the management of an Information System (IS) that has become complicated by the accumulation of multiple layers of applications. Above all, it relieves employees of repetitive, time-consuming tasks that ultimately provide little or no added value.
The best-known application is the chatbot. More and more companies are deploying them, or plan to do so in the near future. A study by Gartner estimated that these virtual advisors would power 85% of all customer service interactions by 2020.
Prioritize repetitive tasks
But beyond these virtual assistants, what method should be applied to ensure the success of an RPA-based project? A company wishing to embark on automation must start by identifying all the repetitive, manual steps performed by employees. The next step is to prioritize them.
But generally speaking, many use cases revolve around the collection, processing, consolidation and execution of data analysis reports. Let's take the example of an order process.
Most requests come from e-mails, PDF files or even faxes. This data needs to be harmonized before being submitted to the back office. And yet, employees spend an enormous amount of time (too much) on data entry and error handling!
In this case, robotic automation of ordering processes involves configuring software - or a "robot" - to emulate and integrate the actions performed by these employees. - to emulate and integrate the actions performed by these employees.
By interacting with applications as a human would, these "robots" can perform even more tasks: open attachments, fill in electronic forms, save and re-enter data, extract structured and semi-structured data from documents...
Reinforcing your safety
It's a mistake to think that RPA is the solution to all ills, and that it will cut costs. Automation should not be synonymous with haste. Automation needs time to gain momentum. Only when it reaches a cruising speed without obstacles can it claim to be "profitable".
On the other hand, despite the obvious net benefits, companies are reluctant to use cloud technologies to automate their legacy IT operations because of concerns about cybersecurity.
By integrating the cloud into its IS and multiplying automation processes, companies absolutely must adapt their IT security policy. On the one hand, cloud automation can help organizations to be much more agile and improve their security.
But it can also extend their attack surface, by creating new risks and loopholes in an increasingly dynamic and complex environment.
So it's important to understand these specific vulnerabilities and respond appropriately.
