Home Case Studies Clerk of the Superior Court in Maricopa County How a spark of innovation developed into a virtual assistant named Cleo
The Clerk of the Superior Court in Maricopa County integrates AI and IBM watsonx Assistant into workflows, improving public services to its team and citizens
Woman smiling with headset on
We’ve all been there. You need to get ahold of a customer service representative for an answer to a quick question. You call and brace to be routed to automated call systems. You endure endless hold times, and you grow increasingly frustrated as you explain your problem again and again to multiple people—only to find out that no one can help.

Many organizations receive thousands of service inquiries a day, which is challenging enough. But add a global pandemic into the mix, and juggling basic call center requests becomes even more demanding. The Clerk of the Superior Court in Maricopa County was experiencing just that.

Based out of Phoenix, Arizona, the Clerk’s Office provides records management and financial services for the county justice system, the legal community and the public. Citizens of Maricopa County, a community of 27 cities and towns and nearly 4.6 million people, reach out with time-sensitive requests, marriage licensing applications, court record access and passport renewal requests. It’s essential that Maricopa County employees can provide citizens with fair and timely access to such services and that such requests aren’t stalled—even if employees aren’t physically in the office 24x7.

Yet handling such large volumes of requests left staff pausing on complex day-to-day tasks to spend most of their time answering calls, addressing routine questions and transferring customers to different departments. As increased handling times continued to mount, so did the frustration of customers and employees. The goal was to centralize the customer experience by applying new technology to provide customers with a better experience. Aaron Judy, Chief of Innovation and AI at the Clerk’s Office, led the office’s modernizing charge.

95.78% Containment Rate


IBM watsonx Assistant held 15,569 conversations with a 95.78% containment rate in 2023

214 hours


IBM watsonx Assistant’s handling of inquiries currently saves 214 human hours per month

When everybody saw the power of a chat agent, the question became, ‘Could we positively impact customer service?’ Aaron Judy Chief of Innovation and AI Clerk of the Superior Court in Maricopa County
Scaling up with Conversational AI

In his previous role, Judy worked in the Office of Enterprise Technology (OET), where he began to tinker with developing a chatbot-based intelligence system. Not much later, he was elbows-deep into researching chatbots, and he started taking build courses through edX—an open online course provider—as a side hobby. He then decided to incorporate his hobby into work by introducing ChatOps.

“I was on a geographically dispersed team that unified through remote meetings and asynchronous communications. We were using Slack to communicate. This was an opportunity for ChatOps to shine,” explains Judy. “So, we started with a Slack chatbot, but at a certain point, my coworkers asked if we could make it enterprise. I realized my piddly little tinkering wasn’t going to scale up to what I needed.”

Judy changed departments and took on a new role as Chief of Innovation and AI at the Clerk of the Superior Court but was asked to continue building a service agent for the office. He started looking at ways to automate actions through conversation. At the time, the Clerk’s Office was fielding 33,000 calls a month, and its primary goal was to diminish incoming call volumes and increase customer satisfaction.

“We needed something with natural language that could understand what you’re saying, extract the important features and act on it. We created a chat agent that could, if you received an alert, allow you to roll traffic from one data center to the other with just a message in Slack,” says Judy. “When everybody saw the power of what that could do, the question became, ‘Could we positively impact customer service? Could we allow users to self-serve themselves somewhere from the service desk?’”

Judy recognized that the Clerk’s Office needed an intelligent, conversation-based, self-service solution that integrated within the existing infrastructure to support its customers and employees across multiple communications channels. Then he remembered taking an IBM edX course on creating an IBM conversational solution, much like a service desk bot.

“One of our design parameters is a solution must survive its creator, and for me to do that, I needed to find something that evolves. I couldn’t just give the product to a layperson and say, ‘Here, care for and feed this intelligent agent,’” recalls Judy. “It just wasn’t possible until I found IBM watsonx Assistant. With IBM technology, we saw how easy it was to transition this tool to somebody that’s not heavily technically inclined; it doesn’t take an expert to keep it going.

I ironed out the care and feeding of watsonx Assistant because it was very easy to transition the technology to a non-technical individual. I said, ‘Here’s my money; let’s go!’ Aaron Judy Chief of Innovation and AI Clerk of the Superior Court in Maricopa County
IBM watsonx Assistant lends a virtual hand just in time

Using IBM watsonx Assistant and Twilio Flex, which is cloud-based contact center software, the Clerk’s Office built a unified omnichannel contact center solution and brought it live within three months. The AI solution, which the team named Cleo, works alongside human agents to provide a more effective support experience for Maricopa County citizens. It uses natural language through omnichannel communications, including phone, web chat, text, social media channels and email. It can also be combined with voice technologies such as Alexa and Google.

By the time COVID-19 reared its head, the Clerk’s Office—with the insights it had previously gathered—was able to determine what services to provide citizens remotely and what actions to feature on its website.

“We actually launched as the office was locking down. They were telling us to stay home; lockdowns were happening,” recalls Judy. “So, I said, ‘Let’s be brave and launch because if nothing else, it will become a more important tool as we are all trapped in our homes.’ And we did just that.”

Thanks to Cleo’s sophistication, Judy and his team can make changes and pivot quickly in real time. The virtual agent is equipped with 151 intents and 3,500 nodes with what Judy describes as “umpteen thousand entities in the middle” that are organically collected from employees working within the courts.

“I started with a PDF that’s given to all employees from the courts, a Q&A worksheet of what you’re allowed to say to a customer,” explains Judy. “Because we are part of the court system, it cannot appear as if we’re giving legal advice. I can’t make anything biased. We also didn’t want just a frequently asked questions tool. Cleo is plugged-in to provide services. You can ask for directions to a facility and it understands where you’re coming from in town and suggests locations with shorter wait times.”

A few years ago, AI in government was still a scary thing, but Watson helped me build that trust. People knew Watson; they saw its capabilities on Jeopardy. I leveraged that social trust in the platform. Aaron Judy Chief of Innovation and AI Clerk of the Superior Court in Maricopa County
No dead ends

The positive results speak for themselves. So far, in 2023 alone, Cleo, powered by IBM watsonx Assistant, has held 15,569 conversations with a 95.78% containment rate, saving agents over 100 hours of call time. Cleo can answer most customer-facing questions, allowing agents to work on other things without having to press pause on tasks to answer the phone.

“Our ability to disperse and disseminate information was pivotal in that, especially with new policies, we were accepting things by mail that wasn’t previously an option,” explains Judy. “Customers also got answers quickly tailored to their specific needs in real time.”

“Part of our design parameters were no dead ends,” adds Judy. “If the customer had to start over just calling the main number to try to get help, that was automatically a bad experience to start. Everyone was really enamored by the potential of watsonx Assistant and Cleo answering these questions.”

If, by chance, someone starts with Cleo and it can’t help, it transfers the customer to the service center for an agent to pick up and continue the conversation. No more having to start over, as the solution saves the history of the conversation for a smooth handoff to a human.

As for the future and looking at the big picture, Judy and the team will continue their relationship with IBM and will work to expand on their audience reach, bolster their channel strategies and refine the Cleo experience. Additionally, the Clerk’s Office wants to ramp up its partnership with human resources by incorporating a curated instance of IBM watsonx Assistant into that area. 

“Cleo opened up many new channels to service citizen questions we were not addressing via the phone channel. We discovered that the phone channel fields distinct engagements that the text channels do not,” says Judy. “So rather than impact those users, we increased access to justice for many more. We are now refining Cleo’s voice experience to allow her to respond to questions via the voice channel. We are and will continue to do a lot of cool things.”

Clerk of the Superior Court logo
About Clerk of the Superior Court in Maricopa County

Located in Maricopa County, Arizona, the Clerk of the Superior Court (link resides outside of ibm.com) is focused on providing citizens and employees with progressive and efficient court-related records management and financial services. Its focus is for the justice system, legal community and public so they have fair and timely access to accurate court records and services.

IBM watsonx Assistant

Deliver consistent and intelligent customer care across all channels and touchpoints with conversational AI.

Lean more about IBM watsonx Assistant View more case studies Camping World

Driving a reimagined customer experience with an AI-powered virtual assistant

Read the case study

Just ask Anna

Read the case study
VIA Metropolitan Transit

Digital agent Ava guides San Antonio transit riders

Read the case study

© Copyright IBM Corporation 2023. IBM Corporation, New Orchard Road, Armonk, NY 10504

Produced in United States, September 2023.

IBM, the IBM logo, and IBM Watson are trademarks or registered trademarks of International Business Machines Corporation, in the United States and/or other countries. Other product and service names might be trademarks of IBM or other companies. A current list of IBM trademarks is available on ibm.com/trademark.

This document is current as of the initial date of publication and may be changed by IBM at any time. Not all offerings are available in every country in which IBM operates.

All client examples cited or described are presented as illustrations of the manner in which some clients have used IBM products and the results they may have achieved. Actual environmental costs and performance characteristics will vary depending on individual client configurations and conditions. Generally expected results cannot be provided as each client’s results will depend entirely on the client’s systems and services ordered. THE INFORMATION IN THIS DOCUMENT IS PROVIDED “AS IS” WITHOUT ANY WARRANTY, EXPRESS OR IMPLIED, INCLUDING WITHOUT ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND ANY WARRANTY OR CONDITION OF NON-INFRINGEMENT. IBM products are warranted according to the terms and conditions of the agreements under which they are provided.