Voice automation support has always had promise.
It's immediate, familiar, and doesn’t require the customer to fumble about with logins or apps.
But it’s been stuck doing the basics for years: routing calls, repeating FAQs, and asking customers to ‘press 1 for billing.’
Now, a new generation of automation tools is pushing voice into more capable territory.
Mavenoid, a company best known for support automation in the hardware and product space, is among those leading the shift.
The company’s Voice Assist product doesn’t just route; it resolves.
“We’ve had people talk to our voice automation system for 40 minutes,” says Gintautas Miliauskas, Mavenoid CEO and Co-Founder.
You don’t stay on a call that long unless it’s actually helping. This isn’t just call deflection; it’s problem-solving.
Voice Is Back, And It's Smarter
Until recently, voice automation had hit something of a ceiling.
Speech-to-text was unreliable, and anything beyond basic menu options tended to fall apart.
Customers would get stuck in loops, grow frustrated, and hammer the zero key to reach a human. More often than not, a few choice words were even thrown in for good measure.
But that’s changing fast.
For Miliauskas, it really boils down to getting two key things right: speech recognition and dialogue handling.
He explained how both had improved dramatically, with word error rates in English currently down to around three percent, and large language models being deployed to help smooth over mistakes.
This has resulted in a voice experience that understands context, handles ambiguity, and responds in natural conversation.
Moreover, it can actually follow through, as the Mavenoid CEO explained:
When you're just running shallow scripts, you can only redirect. But when your system has real domain knowledge, that’s when you start seeing resolution.
From Call Avoidance to Real Resolution
Many automation projects still treat voice as a cost to be minimized.
The goal is to deflect as many tickets as possible, often by encouraging customers to hang up and use a chatbot instead, but that’s not how Mavenoid sees it.
“The opportunity here is to improve quality and reduce cost at the same time,” says Miliauskas.
“You let automation handle the structured, repetitive parts, and your agents are freed up to have the meaningful, complex conversations they’re actually good at.”
And these aren’t empty promises from Miliauskas; his company’s tech is already having an impact for the likes of De'Longhi, Husqvarna, and Stanley Black & Decker.
Another company that Mavenoid is currently working with is electronics brand Broan-NuTone, where Voice Assist has helped to slash average handling time from 10.9 minutes to just two – a 75% reduction.
That kind of efficiency doesn’t just save money; it opens doors.
Miliauskas detailed how the tech is not only helping with cost reduction, but it's also “driving revenue opportunity.
