Automating Nonemergency Calls for PSAPs Using QnABot on AWS with Paragon

Founded in 2004, Paragon is an IT business consulting services firm that helps companies achieve growth, agility, and transformation. The company’s mission is to remove IT department complexity, empowering companies to align their technology strategy with their business goals using proven methodologies. For example, Paragon has worked extensively to bring efficient call center services to its customers through AWS Marketplace, where customers can discover, deploy, and manage software that runs on AWS. Paragon uses AI-enriched solutions to quickly establish an IT department’s capabilities and areas for improvement, with the purpose of finding the most effective path forward. The company uses Amazon Connect, an AI-powered contact center from AWS, to deliver large-scale implementations of its call center offerings in complex environments.

AWS found that public safety answering points (PSAPs)—a type of call center that dispatches first responders, including police, fire departments, and emergency medical services—were facing a shortage of dispatchers alongside attrition rates as high as 25 percent. AWS approached Paragon to help find a solution. “We knew that off-loading and automating those nonemergency calls would be the best offering that we could create,” says Andy Lewis, CEO at Paragon. “It would help alleviate this industry problem across the United States.”

Well over 50 percent of calls fielded by highly trained emergency dispatchers are actually nonemergency calls, such as those relating to animal control issues and local information requests. That mismatch contributed to employee shortages because it led to dispatchers being overextended and unavailable for true emergencies. With its call center expertise, Paragon knew that it was well equipped to take on that problem alongside AWS. “Paragon and AWS began a collaboration effort where both teams brought resources to the table to create the best solution,” says Lewis.

Drawing on its previous experience, Paragon knew that PSAPs needed a reliable, scalable, secure, and affordable solution that would be simple to configure, manage, and deploy—even without an IT team. The company used QnABot on AWS (QnABot) to quickly create capable and compelling conversational AI experiences across multiple channels. Paragon built AgentX, a call automation solution designed for PSAPs. “The availability of a general-purpose solution like QnABot was a major accelerator in our efforts to address this national crisis,” says Adam Mendlik, chief technology officer at Paragon.

QnABot provides multiple mechanisms for generating responses. For common questions, the service can map a caller’s question to a prewritten answer. When the caller’s question is not known in advance, QnABot passes the query into one of several large language models using Amazon Bedrock, a fully managed service for building and scaling generative AI applications with foundation models. Through AgentX, citizens calling nonemergency numbers can receive helpful, generative AI responses to their questions. The solution can even ask callers follow-up questions to deliver tailored responses, automatically create dispatch requests or file police reports, and route a call to an emergency line.

Municipalities can deploy AgentX directly from AWS Marketplace in less than 1 hour. “We designed AgentX to bolt onto the frontend of any existing call center technology,” says Mendlik. “Customers don’t need to tackle a migration project to start using it.”

With the capabilities of QnABot, AgentX helps call centers generate logs with advanced analytics from every channel and view dashboards with near real-time insights into call patterns—all within hours of deployment. As part of the solution, Paragon helps customers harness those insights to identify opportunities for increased automation. In fact, customers can rapidly automate up to 100 percent of their inbound calls, which drastically reduces the number of nonemergency calls that are routed to dispatchers and frees up their attention to focus on emergencies. After a period of listening, learning, and tuning on AgentX, call automation improves rapidly.

When 9-1-1 dispatchers aren’t overwhelmed with nonemergency calls, they can better comply with legally specified performance standards, such as answering calls within 10 seconds. Using AgentX, call centers can answer 100 percent of nonemergency calls on the first ring. Dispatchers can also reduce overtime hours—which comes with the added benefit of lowering costs for the call center. Call centers often budget for up to 30 percent overtime, and this expense can now be significantly reduced or even removed altogether. Overall, call centers can reduce costs by up to 50 percent.

Using QnABot technology, Paragon built a solution that doesn’t require IT expertise and can be deployed quickly. “Once implemented, AgentX pays for itself in the first month for many customers,” says Lewis.

Using AWS Solutions like QnABot, Paragon has empowered PSAPs to rapidly automate nonemergency calls for dispatch centers. That helps minimize the burden on overwhelmed dispatchers and reduce industry challenges across the United States. Already, AgentX is taking off in municipalities nationwide.

Although AgentX was purpose-built for PSAPs, Paragon is introducing the solution to other industries using the flexibility of QnABot.

“At the end of the day, people’s lives are being lost when dispatchers are overworked and emergency calls can’t be answered quickly,” says Lewis. “There are so many positive social impacts of AgentX.”