Enhancing HR Operations at a Canadian Construction Company with Custom Copilot Agent

About the Client

The client is one of the largest industrial construction companies in Toronto. The company manages more than 40 projects across Canada, Northern United States and in the Carribean Islands. The company is driven by a team of more than 1500 employees, working from their headquarters in Toronto and their 40 site offices across North America.

The Challenges Faced During Addressing HR-Related Queries

As they had such a large workforce, the HR department often received a high volume of repetitive queries related to company policies, benefits, leave applications and compliance. Now, these queries were handled by a team of executives. These lead to several challenges:

• Inconsistent responses due to different members of the team answering queries.
• Delayed response times which impacted overall employee productivity
• Inefficiencies in retrieving the information from the right documents

What Made Microsoft Copilot Studio the Ideal Solution for their Challenges?

After extensively engaging with their stakeholders, the team at Evolvous were able to identify Microsoft Copilot Studio. We developed a custom AI agent based on Copilot Studio that used the power of artificial intelligence, machine learning and natural language processing in understanding human queries and responding to queries.

Microsoft Copilot Studio made it easier to deploy an AI chatbot to the human resources portal. Apart from that it offered the following advantages:

• Since it uses a singular AI model, the responses will be more consistent.
• By tapping into the potential of AI, responses were delivered faster and with higher accuracy.
• The AI agent was integrated into the system. It would be able to retrieve the required data from the internal repository quickly and efficiently.

Therefore, we worked on developing a customized AI agent with the help of Microsoft Copilot Studio to enhance the HR-related query experience.

How Did the Custom AI Agent Work?

The custom agent we developed with Copilot and deployed to the HR portal for the company. Here is the workflow of the custom AI agent.

• The client logs in to the system, and then to the AI copilot chatbot using their company credentials.
• The user enters the question to the AI agent. The AI agent answers the queries based on the information already fed to it. If required, it will ask follow-up questions to the user for better context.
• If the query so demands, the AI chatbot will retrieve the required document and then deliver it to the user.
• If the AI chatbot is unable to resolve the query by itself, it escalates it to the relevant HR executive.
• Once the interaction is completed, the Copilot agent logs it, along with the final status, to the company database. It also sends an automatic notification to the client.

The Process We Followed to Implement the Solution

To minimize the time for deployment, we followed a streamlined process. Here is an overview of the process that we followed to develop the custom Copilot agent:

• Our team engaged in extensive discussions with the client to understand its needs and its HR processes.
• After identifying Copilot Studio to be the ideal solution, we started to procure the necessary licenses and set up the environment.
• The team started developing the custom Copilot agent from the ground up
• Once the Copilot agent was ready, we trained it with the relevant documents and company policies. This ensured that the system aligned with their needs more accurately.
• We formed a test group, by bringing together team-members from different departments and hierarchies. They engaged with the chatbot, and we gathered feedback from their interactions. Any issues that were detected were immediately resolved and certain improvements were made
• Once the system was tested thoroughly, we deployed the custom agent to the live environment, where teams from across the organization could use it.
• The AI chatbot’s performance was analyzed and continuous improvements were made to the system.

The Results We Achieved with the Customized Copilot Agent

Once the Copilot agent was deployed, HR queries were answered more quickly and accurately. It was also successful in reducing the overall workload of the HR department. Here are some of the key results that we were able to achieve:

• 90% reduction in HR workload by automating repetitive queries.
• Significant improvements in the consistency and accuracy of the responses, eliminating ambiguity.
• More than 95% faster response times to HR-related queries.
• Seamless and secure access to relevant documents.