The days of manually filling in documents, in-person verifications, and email communication are behind us. More businesses are turning to robotic process automation systems, advancing their customer experiences and in turn, improving business efficiency.
The Case of Patrick
Patrick works in a motor vehicle insurance company. Like any insurance business, they are driven by photos and scans of forms … thousands of them. Tens of thousands. Patrick is drowning in paper and photos – and yes, often the wrong photos.
To make matters worse, Patrick’s director wants him to drive down costs and improve the customer experience – and he thinks RPA technology is a great way to achieve that. But Patrick is worried. Just looking at the onboarding process for new customers, there are some great candidates for RPA automation: capturing details into the back-end system and creating the files. Patrick thinks, with a lot of work, he can bring this process down to a few days. But he knows what his director will say: “I can sign up to Amazon in seconds – days won’t cut it.”
The key problem is that it still takes a human to perform certain tasks, which requires unique visual and cognitive skills to look at and interpret documents.
Certainly, somebody needs to look at each ID document, read the claim forms, and confirm the photos are what they say they are. Often, a photo is simply the wrong type; for example, their policy for onboarding requires an ID document, not a passport but customers often send the wrong photo
These issues are only discovered late in the process. Claims are also mailed to a department, and again only after the emails are read, is this discovered, which slows down the entire process.
These tasks still need to be performed by a human being, which means teams of people, queues, and delays.
Or do they?
Computer Vision and Cognitive Services have made significant progress in the last few years and represent the key to keeping businesses happy.
Mary, the internal development team lead found out about Vision Docs and Vision Verify. This might be what Patrick needs: Vision Verify can automate the verification and onboarding of new customers – automatically validating images of ID documents against Home Affairs, as well as using face recognition and liveness detection to determine, comprehensively, that someone is who they say they are – within 45 seconds to a minute.
Vision Docs will correctly identify the documents, extract the data, and create events that can be used to populate the case in CRM automatically. All document scans are then fully and instantly searchable based on text within the images.
After getting a demo from Mary, Patrick sees the potential and signs up for a trial.
The first thing he tries is the customer onboarding process. Mary wires the sign-up form to send the customer’s name and email (or phone number) to the Vision Verify service.
The Vision AI Customer Experience
Patrick loads their website’s customer sign-up form and enters his details. His phone instantly buzzed with a verification request email. After clicking the link, Patrick was requested to snap a photo of his ID. The next step used advanced AI and face recognition/detection, where his phone required he move his head slightly, to confirm liveness while checking his face against the ID image and home affairs. Finally, he was asked to confirm a piece of information – in this case, his home address. After this step, he checked in CRM and the validated customer tick shining next to his name. In under a minute, he had been validated!
AI Computer Vision for RPA
Computer vision is the critical piece in an RPA initiative. Documents can be classified, extracted, and validated against financial and government databases. Faces can be compared, recognized, and checked for liveness. Humans are not completely out of the equation but only have to deal with the complex edge cases that automation has flagged as being problematic.