“..But what can AI do for me?”
It’s the question we hear time and time again. We hear this from clients, colleagues, industry peers, and even friends and family. AI is everywhere right now, but for business processes, how can it actually add value?
The truth is, AI’s impact depends entirely on how it’s applied. The real question isn’t just what AI can do, it’s what you need it to do. Starting with the problem you’re trying to solve is key to understanding whether AI can genuinely improve your processes.
We’ve been working with clients to explore exactly that, using the Copilot Studio platform to build AI-powered agents that help streamline everyday tasks.
In this blog, we’re sharing a recent project where a client came to us with a clear challenge: they wanted to improve how they respond to Request for Proposals (RFPs) using AI.
Streamlining RFPs with a Copilot Agent
If your organisation relies on RFPs, you know how resource-intensive they can be. Tight deadlines, repeated questions and the pressure to find the right content quickly all add up. It’s a prime example of where AI, especially a tailored Copilot agent, can deliver real value by saving time, improving consistency and reducing the manual effort required to get responses out the door.
A client came to us with a specific challenge: their team was spending too much time manually searching for past proposals and CVs to include in RFP submissions. They wanted a Copilot agent embedded in Word that could help streamline this process by surfacing relevant content directly within their working document. The agent needed to draw on their internal IP, including previous proposals and CVs, as the knowledge base.
To set the project up for success, the client had already done a great job preparing their content. All documents were uploaded into a central SharePoint library and tagged with relevant metadata—laying a solid foundation for AI to do its job.
When designing the solution, we considered two key data sources: the SharePoint document library and Azure OpenAI. Our goal was to test both, to see which would return responses ready for immediate use and which might require additional editing.
In the end, we used both sources within the agent. The responses from each were logged to a spreadsheet, capturing the question asked and the output from each source. This allowed us to compare response quality and make informed improvements.
Behind the scenes, a Logic App kept the SharePoint library in sync with Azure Blob Storage. Every file was tagged with metadata, including the original SharePoint URL. This ensured that employees using the agent could easily access the original source document, without needing to navigate blob storage. The metadata also played a crucial role in refining search results, helping users filter and drill down when needed.
Working within Azure OpenAI Studio gave us additional control. We were able to:
- Choose the most suitable AI model,
- Fine-tune indexing and search parameters,
- Optimise responses to minimise editing time.
The result? A tailored Copilot agent with the potential to save employees hours of manual searching. Now, instead of digging through thousands of documents, they can simply ask the agent a question directly in Word and receive a context-rich response ready to drop into their RFP.
Copilot agent workflow in Word: retrieving relevant content from SharePoint or Azure sources to help draft and refine RFP documents.
Example of a Copilot agent embedded in Microsoft Word, allowing users to access and reference content from connected sources, like SharePoint or Azure, without leaving the document.
So, where should I start my AI journey?
If you’re thinking about using AI to improve a business process, start with one simple question: what problem are we trying to solve?
Look at the tasks that take up too much of your team’s time. Which processes feel repetitive or slow? What’s something you’ve always done a certain way, simply because “that’s how it’s always been”? Do your frontline workers need faster access to information?
Rolling out Microsoft Copilot is a strong first step. It can deliver immediate productivity gains by helping people write, summarise, find and organise information more efficiently, right within the tools they already use.
But to really address the unique challenges in your business, that’s where custom AI-powered agents come in. These can be designed around your specific pain points, whether it’s searching across your internal knowledge base, generating tailored content, responding to repeat queries or automating manual workflows.
Talk to your team. What slows them down? Where are they spending time that could be better used elsewhere?
That’s where your AI journey should begin, and where it has the power to deliver real impact.
About the author
Katherine Collins is a certified business applications consultant at Engage Squared, and has spent the last few years working closely with companies in a few different industries (FMCG, non-for-profit and consulting) to help them get the most out of SharePoint and the Power Platform. Katherine loves helping to streamline the internal work processes and make people’s jobs easier. Katherine is based in Melbourne with her Samoyed and on the weekend, you can find them both in the garden (Katherine working and her dog supervising).
Like what you see?
We’re experts in Microsoft Copilot Studio, helping organisations design AI-powered solutions that transform business processes. Whether it’s streamlining how you respond to RFPs or rethinking repetitive workflows, we’ll work with you to design the right solution.
If you’re ready to explore how AI can transform your business processes, get in touch with us today!
