How Food Cycle Science brought phones in-house and leveled up service with AI Voice Agent
It was pure determination that led Andrew Bird, Head of Customer Service and Experience at Food Cycle Science, to rethink how his team could deliver fast, helpful support at scale.
Food Cycle Science is a Canadian-based clean tech company that makes FoodCycler, an indoor food waste recycler. These appliances reduce food waste by drying and grinding scraps, often cutting volume by up to 90 percent. Less methane gas is released into our atmosphere compared to sending these scraps to landfill, and users make a positive environmental impact from their very own countertops. Currently, the company serves customers across North America, and over 20 countries worldwide through local distribution partners.
“I am fixated on owning the entire customer experience…the motivation was to bring the phones in-house, and I certainly did my due diligence.”
The challenge
When Andrew joined Food Cycle Science, calls about FoodCycler support were routed to a third-party answering service that took messages and created tickets. Customers who phoned in expected answers to their questions and concerns, but were often told they'd get a call back or a note instead. “We were paying quite a bit of money for someone to answer the phone, take notes, and create a support ticket,” Andrew said. “It was not a great experience for a customer expecting immediate help.”
The first step in transforming their phone service was to bring phones in-house. The second was finding a way to extend the team’s reach without losing human touch since many of FoodCycler’s callers are seniors who prefer to speak to someone and often need extra time. With a small team of three, coverage limited to Eastern business hours, and long-running conversations, customers calling later in the day (especially on the West Coast) risked slower responses and an inconsistent service experience.
The turning point
Andrew evaluated several vendors and tapped into a customer success peer group for recommendations. After researching Aircall’s solutions, he made the decision to introduce AI Voice Agent from day one. The goals were clear and simple: to improve the caller experience, capture rich context upfront, and give the team a head start on every case.
The solution
FoodCycler’s AI Voice Agent, also known as Katie, was designed to meet callers where they are and to gather the right information the first time. Andrew and his team set up dedicated agents for business hours, weekends, and holidays, ensuring that callers always received clear guidance and next steps. “If it’s a holiday, Katie explains that our office is closed, then starts diagnosing the issue and collecting details,” Andrew says.
Additionally, Katie is trained to triage, not just take, messages. Callers are asked to provide their contact details, their FoodCycler model, the issue they’re experiencing, and the troubleshooting steps they’ve already tried. Katie confirms details out loud and asks for corrections. This exchange reduces back-and-forth and avoids repeating steps that a customer completed on their own.
Andrew invested his own time to test, refine, and extend Katie’s knowledge. “I sat at my desk for hours, asked questions in different ways, then tweaked Katie’s Knowledge Base,” he says. “If someone needs instructions, Katie asks whether they want to write things down, then pauses between steps. That has helped a lot of customers.”
Aircall’s integration with HubSpot ties it all together. Every call generates a ticket in HubSpot, complete with the call transcript and an AI-generated summary, so agents have context from the moment they open the ticket and can respond more quickly.
Flexibility matters too. Because Aircall is a software solution, Andrew’s team can serve customers from the office or at home without any trade-offs. Additionally, due to the lack of a receptionist, the mobile app acts as a good backup for their Office Manager. Voicemail (including transcripts) also ensures that messages reach the right person quickly, even if someone has momentarily stepped away from their desk.
“By capturing so much information up front, we don’t have to spend 20 minutes on the phone gathering details for each case. [AI Voice Agent] expedites everything.”
The results
Within weeks, Katie became a reliable teammate for Andrew and his small staff of service agents. Overnight calls no longer create a morning scramble. Instead, the team starts each day with ready-to-work tickets, complete with caller details, issue summaries, and in some cases, new responses from customers they’ve been interacting with for a number of days.
“We can clear the new overnight queue pretty quickly,” Andrew says. “Within 20 to 30 minutes between the three of us, we can respond and move cases forward.” That head start helps the team stay focused on follow-ups, resolve complex cases more efficiently, and promptly address incoming tickets or calls during business hours.
“The addition of AI Voice Agent as well as Aircall as a whole has drastically reduced the time it takes for us to provide a first 'human' response to a customer: We went from an average of 29 hours in 2025 to 12 hours by January 2026.”
Agents also spend less time collecting basic information, which makes every interaction more valuable. Customers feel heard, and agents can go straight to a deeper diagnosis or the next step. Katie’s ability to read back information and confirm accuracy reduces miscommunication and speeds resolution.
"I believe that at least 60% of our clients, maybe closer to 70%, are willing to engage with our AIVA, Katie, when they call in as opposed to immediately requesting a live rep. I recently took a call from a customer who said they were pleasantly surprised at how easy it was to converse with Katie and that the empathy and attempts to assist were very satisfying."
The experience for callers improved after hours, weekends, and holidays as well. Rather than a static voicemail, they now receive helpful guidance, relevant instructions, and a clear path to a human follow-up when needed. In some cases, Katie resolves simple questions without human intervention, further easing the team’s workload.
"From a qualitative standpoint, the transparency of informing customers when they can get help from our live team has been very useful when they call during our off hours."
Finally, the overall setup gives FoodCycler’s customer service team the coverage and portability they need. “It’s seamless,” Andrew says. “Wherever our laptops and headsets are, we can serve customers.”
Key outcomes:
78% reduction in time to first human response
4.5x faster response times compared to peak levels
60 to 70% of callers willingly engage with the AI Voice Agent instead of immediately requesting a live agent.
Looking ahead
Andrew continues to research AI and how it can further support FoodCycler’s customers, staying up to date with Aircall’s latest release notes to see if there’s anything else he can adopt. He also makes sure to add new information to Katie’s Knowledge Base as product questions evolve, such as end-of-life policies for legacy FoodCycler models. In terms of future goals, he’s eager to explore more granular routing driven by caller intent, deeper field updates in HubSpot, and smarter suggestions when Katie detects a knowledge gap. “I will definitely be an early adopter when I see that,” he says.
As Food Cycle Science expands into new markets and develops its B2B division, Aircall will help the team maintain high service levels without adding unnecessary complexity. The goal remains the same: give customers fast answers, respect their time, and make every conversation count.
Advice to others
Andrew’s guidance is sincere and driven by perspective. After having a negative experience with customer support AI as a caller himself, he had very important notes to take away and learn from.
“You must put yourself in the shoes of the customer approaching your system. Look at the different situations they face and how quickly they can get to what they need…If they can get there efficiently while still in a good mood, then you've found the sweet spot.”
He also advises leaders to not only invest time into AI upfront, but test and tune the experience often, especially if there’s concern about the quality of the service AI is offering. “With Aircall’s AI Voice Agent, it’s been easy to direct and fine tune the agent so it provides the best quality advice.”
Ready to see what AI Voice Agent could do for your team? Explore how AI Voice Agent can help you deliver faster support, capture richer context, and free your team to focus on what matters.