Automation and robotisation are nothing new. But with the help of AI, they are truly starting to make their mark on all professions and organisations, from the smallest to the giants. This is where RPA technology, Robotic Process Automation, comes into play.
When you hear “RPA”, AI is never far behind
In RPA, the tasks of a human operator are recorded in order to be reproduced automatically. Alexander Levy, AI and Smart Automation Manager at Talan Consulting, and Co-Facilitator of the webinar (in French) “Succeeding in your RPA project” goes so far as to say: “RPA should be considered a digital employee”. An employee ready to take on all the least interesting tasks...
When RPA is in the picture, artificial intelligence is never far behind. It is based on rules, as is often the case with AI applications, enabling the execution of very complex tasks. The combination of the two makes for a very fruitful mix. For example, AI will identify the data types in a text (surname, first name, address, etc.), while RPA will report this information to the appropriate categories. With Machine Learning, AI can even make predictions based on data history. This is known as smart automation, which can be combined with RPA.
Optimising certain activities
Clearly, RPA offers many advantages. Firstly, it helps save time by handling all the repetitive and time-consuming actions with little added value. It is also more capable of adapting to load peaks and troughs. Concurrently, it takes tedious tasks off users’ shoulders, so they can dedicate themselves to tasks with higher added value. Moreover, given that human errors due to fatigue or strain are de facto minimised, the company collects higher-quality information in its databases.
However, not all activities are suited for RPA. As mentioned above, only those with tasks based on specific rules are eligible. Examples include the extraction of data from emails, PDF documents or forms, the retrieval of email attachments or the automatic mining of specific information from the web. Not to mention high-speed comparison between data from different sources...
How to make your RPA project a success
As a result, if you set up an RPA project, one of your first steps will be to identify potential candidate tasks in your company or organisation. In the current context, since RPA is still little-known, it raises questions. “To set out in each of the professions, a gradual ramp-up is in order, with three stages: demystifying, raising awareness, and engaging” explains Jihad Tafroute, AI and Smart Automation Manager at Talan Consulting and co-facilitator of the webinar. In a sense, these are the three keys that open up a project’s doors.
- TAKING AWAY THE MYSTIQUE
This stage is all about explaining to the business divisions what RPA is, and what it can actually do – or not do. “It is a way of removing the obstacles to uptake. At this stage, you can begin to look into what the possible deliverables could be.”
- RAISING AWARENESS
The request can sometimes be voiced spontaneously, coming directly from your customers, through forms or websites. In most cases, a preliminary stage, with ideation seminars, for example, is needed. The objective will be two-fold: raising the business’ awareness about RPA and its benefits, and developing priority use cases.
Other tools include: process auditing, in which the reasons for low or declining performance are identified, or RPA ambassadors appointed to liaise with Management. Topic-specific interviews can also be conducted with a particular department or profession.
The next step is the feasibility study: if General Management approves the projects proposed as meeting specific needs, and if they appear feasible, the first pilots can be initiated.
- ENGAGING
Once the business divisions have been made aware and the contacts able to carry out RPA projects have been identified, pilot studies can be initiated. Take, for example, a simple case, such as automated invoice processing, taking into account all kinds of unexpected events. Team engagement can be achieved through communication on these use cases, highlighting best practices and benefits.
The implementation time of a use case varies depending on the complexity of the process, itself determined by multiple criteria: type of application and data (are they structured or not?), possible scenarios, exceptions, etc.
Depending on the complexity, the development time can range from 1-2 weeks to 4-6 weeks.
Ultimately, RPA applications optimise existing processes to pave the way for major competitiveness gains. “In practice, we see that these applications are well perceived by users when tasks are actually optimised. The ISD, meanwhile welcome this “intrusion” into their private preserve. The reason is simple: RPA is somehow non-invasive. It does not require a redefinition of the IS, let alone the company’s overall organisational system” concludes Alexander Levy.
FOR MORE INFORMATION:
Watch the webinar (in French) “Successful RPA Project”
