How Arbio Reduced Call Center Headcount by 80%

When Arbio scaled past 700 apartments, their support systems didn’t keep up. Thousands of guest messages went unanswered. Visibility was nonexistent. And the property management software they were using, Hostaway, wasn’t built for this level of complexity.

To regain control, Arbio turned to Conduit.

The Challenge

By early 2023, Arbio’s support team was drowning. As they scaled their portfolio to over 700 apartments across Europe, the operational load ballooned. But the tools didn’t keep up.

“We had zero visibility into how we are doing, into how our guests are doing in terms of communication,” said Iuliia Rymar, Head of Operations at Arbio. “We had zero instruments to prioritize one message over another. And we had also no possibility for any automation.”

Arbio was still relying on Hostaway, a property management system built for smaller businesses. It worked fine at low volumes, but once Arbio scaled past 50 properties, the cracks showed. The inbox became unmanageable, with over 3,000 unresolved messages at one point. Every day, team members would jump into a sea of guest inquiries and start responding. They had no clear structure, priority, or way to know what had already been resolved.

“It was just a matter of growth outpacing systems,” Iuliia said. “We didn’t grow the team at the same pace, which would be fine if we had automation. But we didn’t. And the tools we had weren’t made for businesses at our scale.”

The Solution

When Iuliia started evaluating AI tools, Conduit stood out for both product features, but also the partnership.

“I contacted multiple tools, but only Conduit was super responsive and excited about building something together with the feedback that we had,” she said.

From there, the work began, starting with building a scalable knowledge base. Arbio’s existing documentation was fragmented and inconsistent, heavily reliant on manual judgment. With Conduit, they began structuring information in a way the AI could use reliably: defining rules, standardizing language, and filling knowledge gaps.

This new system gave humans more clarity too. Conduit’s redesigned Inbox offered a clean, structured interface for teams to triage issues, while the Conflicts page exposed problems in the knowledge base that previously went unnoticed.

“What I really enjoy lately is the Conflicts page,” Iuliia said. “This is the thing that actually enabled us to grow the automation rate finally to the 60% mark that we always wanted to achieve.”

Conduit helped Arbio structure their operations around what mattered:

• Prioritizing messages by topic and urgency
• Assigning ownership based on domain-specific knowledge
• Monitoring tag frequency to spot systemic issues early
• Standardizing handoffs between AI and human teammates

We managed to get to 60% of the automation rate… and now we send 10,000 to 15,000 messages fully autonomously.

The Impact

With AI now handling the majority of guest conversations, Arbio is no longer hiring additional support staff to match unit growth. If they add 100 apartments, they don’t need 3 more team members. They simply update the knowledge base.

Operational visibility has improved as well. Conduit’s dashboards and tag analytics let Iuliia’s team track what guests are asking about, and how that maps back to issues in the customer journey.

“The dashboard enables us to actually analyze what kind of communication we have and check how healthy this communication is,” she said. “It’s not just about the AI answering.”

This data feeds back into product improvements, knowledge base changes, and sometimes even larger operational projects, like fixing payment flows or changing backend tools.

At one point, Arbio ran into a deeper challenge: confidential information being shared by AI before it should have been. 

“We generally don’t really want AI to know this kind of information,” she said. “What it taught us is that we cannot fully rely on those rules. You can’t just say ‘don’t share until paid.’ You need structural enforcement, not just phrasing.”

That incident led to better safeguards: redirecting to a reservation page with access logic, rather than embedding sensitive details in the knowledge base.

Looking Ahead

Arbio isn’t done. After closing a new investment round in April, they’re entering a new phase of growth, expanding across more cities and adding new lines of business. Conduit remains central to that plan.

“Now I see the results of it already,” Iuliia said. “A much bigger team is working on it, and it’s really nice to see how it went from this first test to something that lets us say, ‘We don’t need to hire another team member if we grow by 100 units.’”

They’re also testing Conduit’s Voice AI, with plans to roll it out for live guest calls.

Looking forward to the further development. What I really want to explore next is voice AI, and see what we can do there.