Uown leasing collected over $500K debt amount and saved $75k by automating collections with Level AI’s Virtual Agent

Uown Leasing serves retailers in more than 46 states through a network of over 450 retail locations that help consumers with challenged credit scores purchase everyday home essentials, such as beds, refrigerators, and televisions, with manageable payment schedules.

Uown establishes customer relationships that last up to 16 months, starting with the customer's need to purchase and ending with the collection of these funds after a set period. The contact center serves as the sole point of contact with the company, employing four out of every five workers. Stateside managers coordinate with offshore team members in Mexico and the Philippines, routing every inbound call through a single central hub.

This centralized structure broke down late at night. The contact center closed at midnight Eastern Time, creating a six-hour calling blackout. Callers at 1 a.m. reached a silent automated recording and a voicemail box. To handle double-digit growth and recover missed late-night payments, Daniel Klein, CEO of Uown Leasing, turned to Level AI to implement its Virtual Agent solution, which manages transactional customer calls 24 hours a day.

"Those were hours where the consumer would reach an IVR, leave a voicemail if they chose to, and then we'd have to try to get them back during the day." — Daniel Klein, CEO, Uown Leasing

Manual dialing inefficiencies and complex time-zone regulations

As the business was growing, the number of calls received at Uown’s contact center increased by 6% over the year, making it difficult to contact past-due customers at scale without hiring dozens of live collections agents. This growth brings other challenges, such as: 

1. Increasing cost of hiring more agents: Daniel had to decide whether to keep expanding the team or use technology to handle the routine workload. It was getting difficult to balance operational costs with maintaining strong customer service.

2. Around-the-clock unavailability: The center ran 8 a.m. to midnight Eastern, which left the West Coast cut off by 9 p.m. their time. A consumer who called at 1 a.m. to make a payment hit an IVR and a voicemail box, then waited for a callback the next day. Those payments went uncaptured, and a borrower who can't reach Uown when they're ready to pay is a borrower at risk of missing their payment date and losing potential leased funds for Uown. Running the human center around the clock would have meant extra headcount and, therefore, the additional cost.

3. Routine requests took a lot of agents’ time: Taking a payment, moving a due date, and quoting an early payoff. These were quick calls that took the agent’s time, leaving less time to resolve complex issues.

4. Spanish coverage costs more than English: Hiring an agent in Mexico is 30% to 35% more expensive than in the Philippines; therefore, Uown keeps their staffing levels lower in Mexico to manage costs, increasing the load on limited agents in Mexico.

All these challenges, along with the rise in inbound calls, pushed Daniel to use technology to handle tasks that didn't require human involvement. But he wasn't looking for just another bot solution. He wanted a partner that was leading the way in AI, one that could do more than handle calls with bots. He needed a solution that could reliably automate the entire process from start to finish.

How Level AI’s Virtual Agent automates collections, compliance, and real-time settlement negotiation

Five9, Uown's contact center platform, pointed Daniel to Level AI, and two things made the decision easy. Level AI was already integrated and certified with Five9, so there was no custom integration to build.

Second, taking payments means handling card and bank account numbers, and Daniel had spent months weighing that risk. When he saw large digital banks on Level AI's customer list, companies that clear a higher regulatory bar than lease-to-own, he had his answer.

"I saw that Level AI is dealing with a few of the largest financial services companies and dealing with their customer's credit card numbers or their bank account numbers, we figured that we're dealing with someone who's already thought the process through." — Daniel Klein, CEO, Uown Leasing

The timeline was his other condition. Daniel signed between Christmas and New Year's and wanted to go live with all the use cases by March 1. The Level AI team delivered on time at every phase, integrated with Five9 without any impact to servicing while going live, and deployed the following six use cases where Level AI’s Virtual Agent took over to

1. Make payments: When customers call to pay an upcoming bill early, add an extra payment, or pay off their loan, the Virtual Agent captures the amount, date, and payment type to process the transaction. Once an early or extra payment clears, the Virtual Agent states the next scheduled due date.
   
   

2. Check balance: The Virtual Agent retrieves the account status from the database to read back the current balance. If the account is current, the Virtual Agent asks the customer if they would like to pay off the loan and transitions them to the payment flow if requested.
   
   

3. Get payoff amount: To resolve outstanding balances, the Virtual Agent calculates and reads back the exact payoff figure. If the customer decides to clear their balance, the Virtual Agent routes them straight into the payment flow to capture the transaction.
   
   

4. Change due date: The Virtual Agent updates billing configurations directly in the database. For customers switching from weekly to bi-weekly pay, the Virtual Agent realigns the payment schedule to match the new cycle.
   
   

5. Promotional expiration: The Virtual Agent retrieves promotional plan parameters from the database to explain the terms and specify the exact expiration date to the customer.
   
   

6. Payments left: The Virtual Agent queries the active contract record to read back the exact number of remaining payments left on the lease.

Level AI’s Virtual Agent handles these tasks reliably. If a customer needs assistance beyond these capabilities or simply prefers to speak with a person, the system seamlessly transfers them to a live representative, ensuring no one hits a dead end. Because the agent executes these tasks with total consistency, Uown has received zero complaints about the automated interactions.

Beyond handling customer requests, Level AI also provides valuable analytics. By reviewing interactions, Daniel identified a simple friction point: customers were completing payments but then asking for an agent just to confirm receipt, primarily because confirmation emails were delayed. To fix this, Uown switched from email to SMS receipts. Now, customers get a text confirmation while they’re still on the line, which has completely eliminated those follow-up requests.

It's pretty rare that we work with another company that feels just like an extension of our team here, and that is worthy of a big thank you from all of us. We are excited to jump into creating more use cases and collaborating on future enhancements that I'm certain will strengthen our teams, together and individually - Denitza Dimitrova, Head of Quality & Compliance

Automated campaigns handled 30% of the inbound calls & saved $75k in agents’ salaries

Uown managed lease growth without hiring additional contact center agents. The Level AI’s Virtual Agent now manages thirty percent of all inbound calls, which covers about 50,000 accounts a month. This automation means the existing team can manage thirty percent more business volume. Uown even reduced its agent team by five percent through natural attrition, with no backfills for departures during the first half of the year. This saved roughly $75,000 in the agent’s salary costs.

The change goes beyond payroll savings. In collections, timing is everything. If a customer cannot pay the moment they have the money, those funds are often spent on other immediate household needs. The Virtual Agent collects around $200,000 a month, totaling over $500,000 so far. Ten percent of those collections come in between midnight and six in the morning, money that used to sit in a voicemail box until the next day.

I've worked on many projects over my 9 years with Uown, and this one was, by far, one of the largest and most fulfilling. The Level team as a whole has been so supportive and so committed to bringing our vision to life, and it's been no small undertaking in such a short amount of time - Denitza Dimitrova, Head of Quality & Compliance

This automation also changed the day-to-day work for the agents. Previously, representatives spent time collecting payments, leaving them exhausted. With the Virtual Agent handling routine payments, agents can spend their time on longer, more complex negotiations. 

Daniel Klein places the financial return at three to four times what Uown pays for the Virtual Agent, compared to the cost of the agents he would have had to hire. The bilingual feature adds further savings. Level AI’s Virtual Agent handles Spanish at the same cost as English, whereas a bilingual agent in Mexico costs 30% to 35% more than an English-speaking agent in the Philippines.

"We love working with the Level AI team because they match the way we serve our own customers. We meet and exceed expectations consistently, and the Level AI team did that for us through the entire project, from scoping to go-live to the post-live changes we asked for." — Daniel Klein, CEO, Uown Leasing