Most mid-market businesses don't have a dedicated AI team. They don't have the internal resources to evaluate every platform, run a proof of concept, and manage a deployment. What they have is a technology partner – and increasingly, the quality of that partner relationship determines whether an AI investment delivers or disappoints.
Rob Oscanyan, VP of Product Marketing at Intermedia, has watched this play out across the market. "There's certainly interest in AI," he says. "But there's a large group that, while they're hesitant, they've heard the value of it and are now looking for practical applications.”
"There's a lot of ambition there, and that ambition needs to meet somebody who can guide them through it."
For businesses navigating AI adoption, the right technology partner can help turn that ambition into action. Here's what to look for.
The First Question Every Good Partner Will Ask You
One of the most common patterns in early AI adoption was a customer telling their partner "I need AI" – and the partner going out and finding something. Oscanyan is direct about the problem with that approach. "When a customer says 'I need AI,' you have to ask: what are you trying to do?" he says. "What does the 30-60-90 look like? What are the results that you want?"
Without that conversation, an AI deployment has no clear direction. A good partner will challenge assumptions, help define the desired outcomes, and establish what success looks like before recommending a solution. It’s the difference between deploying technology for its own sake versus solving a real business problem.
Understanding Your Call Flow Matters More Than the AI
AI that isn't configured to work with how your business handles calls can create a poor experience for your customers. It doesn't matter how capable the underlying model is – if the AI agent doesn't understand your call routing rules, escalation triggers, or how different queues should behave, it's more likely to fall short when it matters most.
Oscanyan draws a direct parallel to what the UC industry has already been through. "We've seen this story before," he says. "Back in the day with communications, people were just going out and buying a patchwork of solutions – a video conferencing solution from here, a chat solution from there. The downsides were more things to manage, communications siloed. We're seeing the exact same story play out again with AI agents." A standalone AI solution that operates separately from the core communications platform may not automatically account for existing call routing rules, escalation procedures, or other established workflows.
"When a call needs to be escalated, many standalone solutions just throw it over the fence and say, 'Well, we're going to set it to this number and hope everything works out,'" he says.
Oscanyan also makes clear that sector-specific understanding isn't always about the AI model itself. "It's more that the partner who's there to help influence that decision. You really want someone that understands what the business is trying to do, and also has an understanding of their current communication workflow. That's actually ground zero."
