Automation of the general insurance policy renewals for all market segments (monthly, quarterly, annually) for a large multinational insurer.
Large insurance companies utilize many commercial channels for acquiring demand. A frequent option is the use of an ecosystem of brokers. Brokers traditionally are companies of all sizes, from home-based one person firms to multi-department corporations. There is a plethora of small size brokers usually focused on the retail segment and working on a local level, i.e. policies for private vehicles, home protection, life and health etc.
Most of the small-size brokers do not have a properly developed IT infrastructure.
The company decided to provide to the small brokers access to an external web site interfaced with the corporate insurance management suite and also alternatively provide to them the option of emailing the renewed policy PDFs.
As a result, the company decided to implement a cost efficient system of processing thousands of PDFs monthly.
An RPA solution was developed to (‘green-path’):
• Confirm the sender of the received emails checking a list of valid recipients.
• Copy the emailed PDFs to a working directory.
• Open the PDFs and extract from the various fields the data.
• Log in to the corporate insurance management suite and input the data there for updating the customer account.
• Archive the emailed PDFs to various folders according to the type of the contract, the customer segment etc.
• Send an acknowledgement email to the broker that the customer profile has been updated.
The RPA solution included also the logic for ('red path'):
• Moving the ‘not correct emails’ to a special folder to be inspected by the operations personnel, i.e. emails from not authorized senders, no PDFs attached, PDFs with the wrong field structure, many files attached etc.
• Transforming the PDF data to the correct format for input to the corporate insurance management suite.
• Correcting simple spelling errors, i.e., names of days and months.
• Handling the most common errors of the corporate insurance management suite that appear on the input from (branching logic).
• Generate a completion & error report at the end of every PDF bulk processing and email them to designated recipients.
Products Used & Processing Volumes
• RPA = UiPath + Local based OCR engine
• Corporate Insurance Management Suite = 3i InfoTech Premia
• Digital Working Env = 10 PCs used for UiPath. PCs in the same area with human employees
• PC OS = MS Windows
• Access Method = Each UiPath instance had its own Company/Active Directory account for security and monitoring
• Max PDF Bulk Size = 800 PDFs
• Max PDFs Monthly = 6000 PDFs
Skills Used – IT
• Project Management (the same person fulfilled the BA role also)
• UiPath Senior Engineer
• 3i InfoTech Premia Senior Engineer
• MS Windows & Active Directory Admins
• Cisco Network Team for providing secure access to various subnets and PDF archiving folders
• Citirx Admins for accessing Premia
• IBM Tivoli Netcool/OMNIbus for monitoring UiPath
Skills Used - Business
• Insurance Business Analyst for authoring the project Specs and the Test Use Cases
• Two multi-year experienced testers from the Retail Insurance Operations team
Prodution Operations Team
• IT: 1 UiPath Admin – Remote role. The first level support role was given to the local MS Windows admins.
• Business: 1 UiPath user. Usually no more than 30% of her/his daily working time was required for starting, ending and monitoring the UiPath bulk operations. Two users were required for covering annual leaves and unavailability. The role was given to the two testers.
Challenges
The company personnel had to adapt during the project elapsed time to a new hybrid human/RPA working model. The employee training had 3 stages:
1 - Awareness. Information about UiPath underlying that it will do the most tedious and repetitive tasks and it will help the company to acquire more business in a cost effective way.
2 - Training for the Insurance Operations Team to check the UiPath reports that the PDFs have been processed successfully or to intervene for fixing the errors.
3 - Training for the senior management to monitor the combined human/RPA operations and output.
