SHARE

Share this post:

Investopedia describes “insurtech” (the term inspired by its commonly known cousin, “fintech”) as the use of technology to create savings and efficiency in the insurance industry. Investopedia also suggests that the insurance industry is ripe for innovation and disruption.

At Grupo Planetun, we know this to be especially true in Brazil. In the Brazilian insurance market, only 30 percent of the automotive market, 10 percent of the housing market, and two percent of cell phones are insured.

Grupo Planetun is an insurtech company in Brazil poised to take advantage of this growth opportunity. We know the big insurance companies we serve need to reduce costs and improve operations, which is why they seek to partner with us.

Innovating the auto inspection process

In 2017, we developed our App Web de Vistoria Prévia, or Preview Web App, that enables image capturing for auto inspections online. When we released the first version of the application, the primary innovation was that the insured individual could take and submit photos rather than needing to drive somewhere or wait for an insurance representative to come to their location.

Today in Brazil the insurance inspection process takes an average of five days, beginning to end. With our application, images can be sent to the insurance company in an average of five-and-half hours. This is a drastic reduction that is speeding overall inspection time.

Despite these gains, we learned by evaluating app use that 30 percent of customer photos submitted were not usable by insurance companies. For example, the photo might be diagonal, cropped incorrectly, or too dark. Or the customer might have submitted a selfie with the vehicle, which cannot be used for inspection.

We knew we needed to address the 30 percent of unusable photos, so we sought a way to provide immediate feedback to customers.

Infusing artificial intelligence into the app

We were introduced to IBM Watson offerings at Think Brasil in 2018. Following that introduction, we began to see how artificial intelligence (AI) could further the capabilities of our auto inspection app with image recognition.

We spent eight weeks with an IBM Garage team in São Paulo to automate our Preview Web App using the open source IBM Cloud Kubernetes service and Watson Visual Recognition on IBM Cloud. Now the solution can confirm or reject customer photos in real time.

The collaboration between the IBM team and our team of developers was crucial. In addition to our enhanced solution, we came away from our engagement with technical knowledge of the IBM Garage methodology for designing and building applications.

Shaking up the insurance market

Aside from the benefits of workflow transformation and user experience improvement, the project with the Garage team helped us reduce app management costs. By reducing the amount of unusable photos shared through the app, our team no longer needs to manually evaluate and flag those submissions.

Additionally, because the new version of Preview Web App is built on microservices and each system has its own API, we are free to offer our customers only what they need.

Through the Garage project, we saw that the agile methodology improved our workflow, so we adopted it internally in our organization as well. We came away from IBM Garage with technical knowledge about Watson, AI tools, image recognition and the Kubernetes database, all of which our developer team is replicating with our other employees.

By partnering with IBM, Grupo Planetun has brought radical change to the Brazilian insurance market. We are the first insurtech company to implement an image recognition methodology for insurance processes in Brazil and throughout South America. This is a major differentiator for our business and is driving company success.

The next step for Preview Web App will be to put the Watson Visual Recognition service to work sorting and pricing the amount of damage a vehicle has suffered in accident situations.

Read the case study for more details.

LEAVE A REPLY

Please enter your comment!
Please enter your name here