We all want to have end-to-end integration of systems, but the reality in most companies is that there are always systems that don’t talk to each other and require their own data entry. Robotic process automation can relieve business users from repetitively keying the same data into multiple systems that don’t interface with each other.
SEE: Hiring Kit: Video Game Programmer (TechRepublic Premium)
RPA is used for other types of repetitive business processes, too. It removes labor-intensive tasks and allows each user to automatically place data onto a user interface so the worker doesn’t have to waste time doing it manually.
Despite these efficiencies, there are inherent challenges with RPA. “If a different user tries to duplicate data or if the user interface changes, a new bot is required to continue the automation process,” said Abel Verweg, product marketing manager at Mendix, a low-code process automation provider.
With hyperautomation, technologies like artificial Intelligence and machine learning can be applied to processes like RPA. They can augment the power and the value of what RPA provides.
“In one case, a European-based bank wanted to open a branch in the United States,” Verweg said. “Numerous documents are required of people who wish to apply for a bank account, and manually checking and validating these documents is a massive time drain. Managers would like to intelligently automate them. This could look like an automated database review. Employees get notified if a customer has no flags in the database or if additional research may be required for various reasons, such as documents that reveal a customer is already a member of the European branch, or miscellaneous transactions may need to be reviewed. In this case, AI tools were brought to the user experience, which was core to what the bank needed.”
SEE: Digital transformation: A CXO’s guide (free PDF) (TechRepublic)
In the project, the users worked with Mendix, which had expertise in automation. But regardless of whom users work with, a strong collaboration between business process experts and technical process enablers is needed.
“With hyperautomation, you typically see IT people who can build a system that tries to work alongside the business,” Verweg said. “When communicating to the business how a process runs, the communication—or more often the miscommunication—is the biggest obstacle. This is where the frustration starts. A business sees an issue and wants to digitize it. Hyperautomation is about being fast, but it’s also important to have a person who understands the process working with someone who can build it. They must collaborate and build a solution together. At this point, it’s not about time-to-market. It’s about time-to-value.”
Let’s see how companies can pragmatically add AI in the form of hyperautomation to an RPA process.
First, an appropriate business use case must be found. An example would be incorporating AI into an RPA process by having the AI remove manual work by scanning an image to establish how many pieces of inventory someone has at a distribution center or scanning through a large number of documents to find a specific word for a law firm.
Next, business users and IT must coordinate the project. “A citizen developer can use a hyperautomation toolkit,” Verweg said. “In our case, we have two integrated development environments. They use the same underlying model, which means you’re not working in a separate environment. As a professional IT developer, you can create custom Java actions or get really technical, if necessary. You can do it all in the same tool that business users use when they model a business process by using visualizations of process steps. As this happens, the developer can ensure security, deployments, etc., are set up correctly. The synchronicity of the team provides better results.”
The goal is to super-charge your workflow automation with AI that can make decisions on worksteps that employees formerly made, while effecting end-to-end process automation.
For CIOs and business leaders, hyperautomation could be a step up for their RPA projects, because hyperautomation can do three key things:
- It can bring AI into everyday tasks.
- It can process big data such as images (a limitation in RPA).
- It offers a uniform development platform for business users and IT.
“If you want to automate something, you need to wed the efforts of business and IT—all on the same platform,” Verweg said. “It’s best to think about the customer experience and journey, the holistic goal, and build from there.”