Background

I spent two co-op terms at the Kitchener Development Centre, an innovation lab for a Canadian insurance company. The office was created for building moonshots, where the corporate head office let teams of student developers play around with novel ideas and come up with use cases on how to transform the company into a digital leader of the insurance industry. We were playing around with automating data processes, using artificial intelligence and chatbots, and creating web apps to make day-to-day lives easier for insurance agents.

Term 1 - Robotic Process Automation and Databases

Robotic Process Automation of Legacy Systems

Problem: The company required a number of legacy systems to be updated with new customer information, but the current process was to do so in a manual fashion. This meant that multiple employees were spending 20+ hours per week manually entering data into these legacy systems in a way that was prone to human errors, and tediously slow.

Solution: I built an automated process using Excel and UIPath to automatically enter new client Information. The process checked for new clients on a daily schedule, and then inputted the data into 3 separate legacy systems. I also ensured that error handling and messaging was built in, in the case that the legacy software was down, passwords had been changed, or if the customer data was incomplete. The creation of this automation was able to save upwards of 100 hours of work for the company every month, and reduced the rate of errors in the company databases.

Insurance Claims Fraud Database

Problem: Individual agents would collect instances of insurance fraud that occurred within a single office, but the information was generally stored in individual Excel sheets on agents computers. No national catalogue of insurance fraud existed, making it difficult to track repeat offenders who may have switched locations or agents.

Solution: We built a No-SQL BLOB (Binary Large Object) database that allowed agents to centralize their files on clients that had been previously flagged for fraud in other regions of the country. The BLOB aspect of the database allowed them to upload files not typical of a database such as videos, photos and audio recordings. This enabled agents to provide in-depth proof of how these clients had acted fraudulently, and led to better interoffice communications.

Term 2 - Alexa Applications and Web Apps

During my second term, I returned as a developer but took on additional roles as a project manager. In this additional role, I created JIRA and Confluence issues and roadmaps, organized meetings between the various departments collaborating on the projects, created designs using Figma, and collaborated with other developers.

Querying Insurance Statistics using Conversational AI

Problem:

Solution:

Building Modern Web Apps for Insurance Definitions