
The client is one of the major insurance firms in British Columbia, specializing in medical insurance. The firm is headquartered in Vancouver, but serves several regions across the province, including the cities of Surrey, Burnaby, Richmond, Abbotsford, as well as Kelowna and Nanaimo. The company aims to deliver better products to its customers and also simplify the overall experience for them.
The client was struggling with an overwhelming volume of documents being submitted by their clients and hospitals. Collecting and analyzing documents like patient invoices, discharge summaries, diagnostic reports, prescriptions, and other supporting materials, was an important part of the claims management process.
The team was manually reviewing the documents and generating summaries which were submitted for further claims processing. This process was time-consuming, error-prone and was putting unnecessary workload on their claims processing team. Therefore, they were looking for an efficient solution for summarizing such critical documents.
After extensive engagement with their team, we were able to identify Microsoft Copilot and Power Automate as ideal platforms whose capabilities could be leveraged to solve their problems.
We realized that they needed a robust AI solution that made analyzing the documents quick, easy and accurate. Therefore, we decided to build a custom AI assistant with the help of Microsoft Copilot Studio. This platform allows the user to execute specific tasks with simple plain text prompts. Therefore, it can help the team to accurately summarize documents, drastically cutting down the time in processing such files.
Power Automate, on the other hand, was identified as the tool which could be used to automate the entire process of document summarization and sending it forward to the next stage of processing. This greatly reduced the time spent in the process and also minimized the errors.
Therefore, we decided to develop an integrated solution that combined the capabilities of Copilot Studio and Power Automate to simplify the overall process.
The solution we designed for the client, worked in the following way:
As soon a document is submitted by the hospital or client, a Power Automate flow is triggered.
Power Automate flow immediately notifies the members that a submission has been made.
The flow also calls the custom Copilot Agent which we built for the client. The Agent uses the power of AI to analyze the documents and create accurate summaries.
Power Automate gets initiated again. In the next stage of the flow, the summary file which was created by Copilot is stored in the respective folders in their SharePoint library.
After the process was completed, the members of the team are notified that a summary has been uploaded and the link to the SharePoint file is mailed.
The client wanted to deploy the solution as quickly as possible. Therefore, to meet that demand, we followed a streamlined process which achieved fast deployment times, without compromising on the quality of work.
Here is an overview of the process that we followed:
Discussion and Planning : We engaged in extensive discussions with the client, especially with the members of the claims processing team. This helped us to understand their process and the challenges they faced. After collecting the information, we planned the solution and the implementation roadmap.
Designing and Training the Copilot Agent : Immediately after the planning stage, we started the work on designing the AI agent with Microsoft Copilot Agent, in alignment with the company policies and procedures. The permissions were configured so that no sensitive information is misused, and all compliance codes are adhered to. We trained the system with the existing documents that the client already had stored in the system.
Developing and Deploying the Power Automate flow : Once the Copilot agent was developed, we built the Power Automate flow. The logic and branches of the flow were developed and then integrated together. The Power Automate flow was then integrated with the existing platforms so to ensure that the system worked seamlessly
Testing and Deploying the System : After Copilot and Power Automate were integrated together, the solution was tested. A few minor issues were rectified and then it was deployed to the live environment.
Client Support : We provided the client with detailed documentation which could be used as ready reference material. Our team also provided training to select members so they can use the new system with confidence. Evolvous also monitored the system for a period of time and all issues were immediately resolved.
Our integrated solution was able to achieve remarkable improvements in various parameters. Here are a few key ones:
80% Reduction in Manual Processing Time : The solution eliminated the need for manually summarizing the documents. This reduced the time taken to complete the summary process.
94% Fewer Errors : The accuracy of the process also increased significantly, as Copilot agents used AI-driven validation and advanced handwriting recognition.
75% Higher Efficiency : Teams are now able to focus on high-value tasks, instead of such repetitive tasks. This increased the efficiency of the entire team.