ClaimGem

Managing and filing insurance claims after major disasters

Comment 1

Inspiration

In the days and months following a natural disaster, nearly everyone needs the same (scarce) resource: professionals capable of composing formal requests for insurance reimbursement.

  • 450,000 insurance claims were filed in the wake of Hurricane Irma.
  • 670,000 claims were sent in after Harvey
  • Many individuals end up waiting months and months before they can pursue reimbursement (or miss the 60 day window!) losing everything
  • One of the team-members lived in an affected area, and saw his own home destroyed, and experienced the arduous process of trying to get paid back for damages (over the 7 month period, his family just ran on credit cards)
  • We sought to build a platform to make this process simpler

What it does

  • Our team built a system to generate professional-grade insurance claims in the wake of a major disaster.
  • To do this, we created an interactive platform that allowed users to create and manage a large reference base of insurable material, enabling claimants to optimize and correct their claims before submission.
  • For verification, we leveraged DocuSign’s Technology to distribute, validate, and track activity on insurance claims.

Functionally, ClaimGem allows a user to enter home features such as their address, the type of damage received, and the specifics of lost items from within their home (defined in greater detail below). It uses this data to generate forms that would be acceptable to a professional adjustor (over the course of the hackathon, we spoke with multiple individuals affiliated with the insurance claims process).

How we built it

  • Used create-react-app to bootstrap quickly off into a React web app, then fleshed out the backend with Firebase Authentication and Firestore, and finally tried experimenting with a serverless Firebase Cloud Functions setup before reverting to a local server for processing DocuSign requests.
  • We used GitHub to collaborate on our code and designed our application in a shared Google Doc

Challenges we ran into

  • Building a giant webapp in 24 hours can be a real struggle, but building an aesthetic one is impossible.
  • We tried to run the DocuSign API from a serverless GCP instance, and we couldn't get it to work. Eventually we just built our own local server. We lost a lot of time there.
  • At the beginning of the hackathon, we realized that our original idea wouldn't work with the hardware available. Therefore, we had to pivot almost immediately, and this took up 2-3 hours as we determined the best course of action.

Accomplishments that we're proud of

  • We were able to build a significant amount of the primary workflow. With the exception of a few bugs, all of the major pages in the core work flow are built, linked to each other, and connected to our database. While the website does not look perfect, we were able to design patterns learned from previous independent projects and build a decently appeasing website.
  • First exposure to Google Cloud Functions
  • We're hoping that it serves a public good and helps people get the relief they need quickly.

What we learned

  • How important reimbursement claims forms and formatting are in insurance claims
  • How attention to detail and security dramatically affect your ability to get a return quickly
  • That both insurance companies and claimants benefit from faster processing times and stronger transparency
  • We learned that the main scoring metric for a successful claims adjustor is claims processd per week--as long as we can ensure that a claim is not fraudulent, the average insurance company is happy to accept it
  • We learned how to leverage DocuSign to disrupt large industries based around forms, security and verification
  • We learned how to collaborate as a team, effectively divying up tasks across team members and then merging them after completion

What's next for ClaimGem

  • Using big data to predict fraudulent claims (minimizing insurance agent's time)
  • Doing better predicting of what items people have in their homes (maximizing claimants returns)
  • Suggesting places nearby to purchase replacement items
  • Pre-disaster inventorying and estimating
  • Working with claims adjusters to streamline the process further
  • Partnering with larger insurance companies to further extend our reach
  • Automating the process of requesting quotes from contractors by leveraging networks like Thumbtack and automated documentation systems like Docusign
  • Using cloud computer vision services to automatically analyze and re-format the information of contractor quote forms
  • We hope to generalize our supporting technology (Docusign, Google Cloud, etc) and streamline the development process. This will let us quickly move into other related areas where archaic form management systems enable disruption for great social good.

Built With

Share this project:

Updates