To solve problem of P&C Insurers so that they can register their damage issue, send invoice for claim processing and get professional assigned to visit their home for damage repair.

What it does

Bot registers claim, gets the invoice from user processes it and sent a confirmation mail to user with details including claim refund status and approved amount, Shows list of professional nearby who can repair damage.

How we built it

2 factor authentication using email and twillio OTP, Built using neo4j API graph which stores user nodes with address, policy details etc. Also used Azure VM to handle backend fulfillment processing. Neo4j Mario user Details

Challenges we ran into

Fetching user data like policy, address details, damage type data from neo4j using API. But we incorporated and made it functional for bot flow.

Accomplishments that we're proud of

2 factor authentication using email and twillio OTP, Using neo4j and Google map API, IDP for invoice image processing, Computer vision model for damage image classification.

What we learned

2 factor authentication using email and twillio OTP, Conversational flow creation and designing, API calling of neo4j graph, IDP for invoice processing, Computer vision model to classify the damage image into type of damage like water, fire etc, map api for displaying and plotting geospatial location using latitude and longitude.

What's next for Virtual Assistant for P&C Insurer

Scaling it to add Live agent transfer to handle claim refund issues etc.

Built With

Share this project: