When a call center needed to scale, Insight US answered with Azure OpenAI

March 17, 2025

At a time when AI adoption has reached its peak and businesses are eager to dive in headfirst, Microsoft partner Insight US believes it’s just as powerful to pause before taking the plunge. That means thinking strategically and tactically about where AI can drive the most value in your business—without getting caught up in a long list of lofty ideas. “You can have lots of great ideas, but if you don't have clean, accessible data, if you don't have the right things in place, then AI is not going to solve anything,” said Anna Chapman, Client Solutions Director at Insight.

But in a tech landscape that’s always changing, how do you balance being thoughtfully strategic and agile enough to keep pace? For Insight, the answer is easy: with the power of partnership.

For more than 30 years, Insight has deployed and delivered Microsoft solutions for organizations across nearly every industry to help them navigate—and stay ahead of—the shifting landscape, all while growing their technical expertise along the way. (They’re an Azure Expert Managed Services Provider, with all six Solutions Partner designations and 23 specializations to show for it.)

Partnering with Microsoft hasn’t just helped Insight grow, though—it's reinforced a deeper level of trust with their customers, too. “Microsoft and [Insight] are walking in together, putting skin in the game to say, ‘We’re here to help you,’” said Chapman. “It builds confidence with customers knowing Microsoft is validating that we are the partners to deliver the type of [solution they need].”

The call center conundrum

The retailer’s call center used interactive voice response (IVR), an automated system that directs callers to relevant information based on menu options they can select by touch-tone or spoken responses. (Think: “For store locations and hours, press one. For order or tracking status, press two. To speak to our team, press three.”) Most of the calls the retailer receives are straightforward questions about store hours, inventory, and billing issues, but they also get more complex questions about in-store appointment scheduling and account information.

Previously, the retailer’s IVR system needed human agents to answer about 90% of those inquiries, so their team spent more time fielding basic questions than focusing on store operations or in-store customers. On the other hand (or line, for that matter), this created a frustrating experience for the customers on the phone—being put on hold, getting transferred to the right store, then waiting for someone to answer. Coming out of the Envisioning workshop, Insight’s team was confident that AI could handle 90% of the responses to these inquiries. Plus, through their partnership with Microsoft, Insight could tap into funding with Azure Innovate—showing the retailer that Microsoft believed in the project, too.

Insight’s first priority was to make sure the retailer’s existing architecture could support the work they planned to do for their call center. According to Insight’s National AI Practice Lead, Meagan Gentry, there are three core layers that set the foundation for an AI solution:

In this case, it turned out that Azure OpenAI was the glue they needed to deploy an interactive voice assistant (IVA) powered by generative AI. With an IVA, callers would be able to instantly get the answers they needed, bypassing the tedious process of selecting a menu option, being put on hold, and waiting to get transferred to the call center or the right store.

Fortunately, Azure was already part of the retailer’s architecture, which made it that much easier to work with their existing data. “There were a lot of different data sets needed to do inventory checks and account lookups,” Gentry said. “The PostgreSQL and Azure SQL databases [the retailer] already had integrated really well with the Azure OpenAI search connectors, so we could just reach in and grab that data.”

The retailer was also already using Azure Communication Services to handle call volume traffic, so they could seamlessly implement Azure AI Services and features like speech-to-text and text-to-speech. “The good news is that all these [products] play nicely together, and that's what makes Azure a lean-to platform for solutions like this. [Azure’s] integration methodology is documented, it’s out there for us to learn, and it’s very low lift,” said Gentry.

From there, Insight tested different paths for the call center’s prompt and response behavior—which involved thinking through questions like, “Is this potential question getting routed to the right place? Is [the response] correct, helpful, and leading to resolution?” They made sure each prompt got to the right place with Azure OpenAI, and with the help of a retrieval augmented generation (RAG) framework, they could further customize the responses based on the retailer’s unique data sets—securely.

“Azure AI is constantly responding to market changes in a safe way,” Gentry said, “and that creates confidence with our customers that when they’re using Azure, they’re going to have a menu of options for AI at their fingertips that meets their needs but is also well-vetted.”

Partnering to turn AI visions into reality

With the IVA, Insight expects 40% of calls to route to a human agent instead of the previous 90%, freeing up their team to focus on higher-value work like improving store operations and engaging with customers. From that first Envisioning workshop through implementation, though, Insight’s team made it clear that their support doesn’t stop at the technology: they operate as part of their customers’ teams, looking ahead at their long-term goals and working side-by-side to reach them.

In this case, Insight has not only continued to give guidance and fine-tune the retailer’s call center solution, but they’ve also been having conversations about their data and AI roadmap and exploring more AI use cases that align with their business goals. “We don't come in with a pre-baked cookie-cutter solution; we work with customers to understand the business outcomes they want to drive,” said Chapman. “Then, using our various tools in our toolbox—whether it's hardware, software, or consulting services—we help them further their vision.”

Together, Insight’s technical expertise and the industry insight, product team support, and funding from Microsoft made the customer’s AI ambitions a reality. “Insight brought their hands-on expertise to get into the weeds of solution building and delivered an AI workshop pushing the leaders to think differently and prioritize and identify impactful use cases,” said Daniel Garney, Senior Modern Workplace and Microsoft 365 Copilot Specialist at Microsoft, who worked with Insight on the project. “Microsoft painted the vision and art of the possible, and Insight brought that vision to life through their services arm,” Said Jessica Milligan, Senior Azure Data and AI specialist at Microsoft.

And with every AI vision that Insight brings to life, they build upon their experience and lessons learned to further hone their skills, strengthen their practice, and scale their impact. “We’re carrying some of the things we learned from this to healthcare for emergency response protocols, or to education spaces where curriculums are being designed with Azure OpenAI,” Gentry said. “I think that’s one of our superpowers with Microsoft: we're learning together, and we're taking those learnings and propagating them in a lot of other places.”