For Flex, AI accuracy isn't a feature, it's the job
Flex helps renters stay on track by letting them split monthly payments and avoid late fees. But rent is more than a transaction — it’s tied to housing security. That means high urgency, high emotion, and a high volume of support tickets a month. Flex began the search after deciding to migrate off Ada, whose one-size fits-all flows couldn’t match the policy precision renters rely on. In a space where timing and rule accuracy directly affect housing stability, guessing was not an option. They needed an AI system that could stay calm under pressure, follow strict rules, and know when to escalate. Flex put Lorikeet in a head-to-head test against Decagon — and Lorikeet won across every metric that mattered. It didn’t just reply fast; it got it right. Trained on Flex’s policies, workflows, and tone, Lorikeet now handles the bulk of high-volume tickets and escalates with context when needed — giving renters clear answers, and Flex peace of mind.
From Day 1, Flex & Lorikeet worked side by side
Integrations are where AI rollouts usually get messy. But in this case? The setup was easy. Lorikeet plugged straight into Flex’s existing API and authentication flow — tapping into 40+ customer attributes across account, transactions, property integration, and feature availability to deliver hyper-personalized responses. From there, Lorikeet ingested SOPs and backend data sources quickly and efficiently. What started as a fast integration became a shared learning loop — one that surfaced content gaps, trimmed duplication, and made the agent smarter with every ticket.
Flex onboarded Lorikeet as if it were a human agent
When rent’s due, every second counts — and so does every word. Flex didn’t just need automation; they needed an AI that could interpret date and mathematical terms to best help their customers in high-stress moments. So the teams tackled edge cases together: building logic to interpret “today” vs. “Friday,” adding disambiguation prompts for vague payment questions, and flagging tickets by urgency or sentiment. The lesson? More information isn’t always better — the right information, delivered clearly, is. Lorikeet was trained like part of their team, with workflows shaped by how Flex actually operated — not just what the SOPs said.
No delays. No dead ends. No loops.
Lorikeet was trained with strict escalation logic to flag high-risk tickets immediately, including urgency signals, emotional language, or complaint sentiment. When confidence dips or discretion is required, Lorikeet steps aside fast — never trapping someone in a loop.
“We’re excited to explore other ways Lorikeet can help us grow Flex — like through email or voice support. That way our human agents can focus on solving more complex customer challenges.”
Lindsay Boland
Customer Service AI Product Lead
When rent’s due, chats spike. Lorikeet handles the surge.
Flex needed an AI solution that could keep up with the pace and pressure of the making a monthly rent payment. With Lorikeet, customers get faster, clearer answers— and it supports Flex’s brand mission of giving people more financial control and peace of mind each month. Now the Flex support team has more time to focus on solving more complex customer challenges, and scaling their business.
