#Insurtech #RPA #Commando
Time To Market : 4 weeks
"I was not familiar with this type of project, I am rather impressed by the deployment speed in view of your limited knowledge of our profession. I appreciate the global view. I did not know it would be possible."
Cindy Piau, Indemnification team Manager
The recourse process between insurers determines the insurance responsible for the compensation of a claim. The processing time of each file took 48 hours. GAN had 7000 files late to process, representing 2 million euros of uncollected compensation.
We developed a robot to automate the treatment process for these claims files.
The key points:
late files to process
per month to recover in compensation
saved the first year
Today, GAN managers manually enter the information received on pre-recorded DAB and auto in their tool, Oasis.
This process is a source of mistakes and time and productivity loss for the GAN managers:
In order to develop the robot in less than 4 weeks, the team connected to an existing tool on a new stack that was adopted in 1 week. Then, to accelerate the scrapping of data, pair programming developments enabled a rapid rise in the skills of internal GAN teams.
We developed two robots triggered by CRONs to scrap the data.
The first robot, the opener, scrapes the data of the claims and automatically opens the file in the manager's software, Oasis.
The second one, the missioner, sends a request for personalized processing to the expert by analyzing the data from an Excel file.
The robot's aim is to allow the managers to focus on missions requiring business expertise, such as assigning the relevant expert for each file, which is the task with the most added value from a business point of view. The challenge was to identify precisely from the very first week the processing steps to automate tasks with little business value.
