The Information Was There All Along

If valuable insights are buried across status updates, spreadsheets, emails, and notes, and creating a clear picture of progress feels harder than it should, keep reading.

When a team is small, staying informed can happen naturally. People know what their colleagues are working on, which projects have moved forward, and where collaboration may be necessary. As the number of people and projects grows, that shared picture becomes harder to maintain. Information is still being captured, but it is harder to see as a whole. 

To address this challenge, the AI Solutions (AIS) team developed an AI-powered workflow that works for them using one AI agent that compiles weekly staff updates while a second agent creates a leadership summary email each week. Staff continue to provide updates as part of their regular workflow, while AI pulls the information together and identifies where there is progress and where there are obstacles. 

THE CHALLENGE

As the AIS team expanded its portfolio of university-wide initiatives, understanding progress became increasingly difficult to manage. Each employee understood the work they were a part of, but creating the executive overview was a challenge. Without a project management solution that fit the team’s workflow, weekly updates were reviewed manually across multiple places, with management trying to identify what had changed and convert each person’s status updates into a standardized format.

“The challenge wasn't getting updates...it was understanding how all of those updates were connected to one another.” 
  - Mark Rosanes, AI Solutions Manager

As the number of projects and team members increased, leadership spent more time gathering and interpreting status updates out of their OneNote notebooks than evaluating progress and addressing obstacles. Different writing styles, varying formats, and the lack of consistent project tracking made it difficult to quickly identify which projects moved forward and where leadership attention was needed. They did not want to change how they were collecting status updates; instead, they wanted to automate the parts of the process that did not require human judgment. 

THE SOLUTION

Using Copilot Studio, AIS created two specialized agents with different jobs rather than asking one agent to do everything.

1. Weekly Status Consolidator (WSC) collects staff updates from OneNote, compares the current week data from the previous weeks, and organizes the information into a consistent structure in SharePoint: project title, status, changes or accomplishments, issues or risks, next steps, and whether manager attention is required.
2. AIS PM Tracker (APT) uses the structured information to create a concise leadership summary delivered to the AI Solutions Manager each Wednesday afternoon. Instead of searching through individual updates, the manager receives one view focused on meaningful movement, risks, and items that may require action.

The agents do not decide whether a project is successful or determine what leadership should do next. They organize and surface the reported information. People remain responsible for interpreting the context and deciding how to respond. 

LESSONS LEARNED

• Having information and having visibility are not the same thing. The AIS team was already documenting its work each week, but understanding project successes and hurdles across dozens of ongoing activities remained difficult. AI became valuable by helping translate existing information into something easier to evaluate and act upon.
• Start with a workflow people already use. Staff were already documenting their work, so asking them to adopt another tool simply to make automation possible would have created a new problem. By working with the existing OneNote process, the solution reduced reporting friction instead of shifting it to someone else.
• Give different AI agents clear responsibilities. Separating the agents’ roles ensured troubleshooting was easier. Rather than asking one agent to perform several different kinds of reasoning, each could be optimized around a clearly defined task.
• Consistent structure improves AI results. Standardizing the reporting format made it easier for the AI to understand changes from week to week.

HOW CAN I APPLY THIS TO MY OWN WORK?

The opportunity for change appears when employees are spending significant time reading across different sources just to understand what has changed or what deserves attention.

A useful starting point is to ask:

• Are people already documenting the information we need, but in different places or formats?
• Does someone repeatedly compare this week, month, or case with the previous one?
• Is leadership spending time assembling information before it can make a decision?
• Could a common structure make changes, risks, exceptions, or priorities easier to see?
• Can AI support the process without forcing people to abandon a workflow that already works? 

THE TAKEAWAY

Many departments already capture the information they need, but the challenge is rarely capturing information. The real challenge is understanding what that information means when viewed as part of a larger whole.

By using two focused AI agents to compare, structure, and summarize information the team was already creating, AIS improved project visibility without adding another reporting system or removing human judgment from the process.

For other teams, the lesson is simple: look at the information you already have and ask whether AI could help people see what matters sooner.