Enhancing public safety operations with Amazon Bedrock

Public safety agencies face challenges including staffing shortages, increasing call volumes, and evolving community needs. One routine task that takes up valuable time for public safety agencies is handling noise complaints, which comprise up to 15 percent of non-emergency calls in urban areas.

Amazon Web Services (AWS) has developed generative AI solutions to address the unique challenges of public safety agencies today. For example, foundation model (FM)-powered agents, offered through Amazon Bedrock, can efficiently manage routine interactions. These Amazon Bedrock Agents process context and nuance, integrating with existing public safety systems to provide functional support, and enhancing agency operations through automation and natural language processing. By using these agents, public safety agencies can optimize resource allocation and focus on critical tasks.

In this post, we will discuss a noise complaint solution that uses an Amazon Bedrock Agent to serve as an intelligent interface between citizens and the computer-aided dispatch (CAD) system. When a noise complaint is received, the agent automatically gathers essential details such as location, type of noise, duration, and severity through natural language interactions. The agent then formats this information according to the CAD system’s requirements and creates a properly structured incident report, while simultaneously providing the complainant with a case number and status update. This automation reduces dispatcher workload, enables consistent data entry, and speeds up response times by eliminating manual processing steps. The agent also maintains communication with the complainant through automated updates and follow-up messages, freeing up public safety resources for more critical emergency responses.

Architecture

Our noise complaint solution engages with users through Amazon Connect, which works with Amazon Lex for natural language understanding. Amazon Bedrock handles complex AI processing, triggering AWS Lambda for business logic execution. The backend uses various AWS services, including Amazon Simple Email Service (Amazon SES) for email, Amazon API Gateway, Amazon Relational Database Service (Amazon RDS), Amazon DynamoDB, Amazon Simple Storage Service (Amazon S3) for storage, and Amazon Simple Notification Service (Amazon SNS) for notifications. The following diagram illustrates the solution architecture.

Figure 1. Architectural diagram of the solution

The generative AI-based solution uses Amazon Bedrock Agents to process natural language, enabling residents to report issues conversationally. For example, when reporting a noise complaint, the solution can extract important details such as contact information, location, nature of the disturbance, previous resolution attempts, and severity indicators. This approach makes reporting more accessible, especially for those who might struggle with traditional paper-based forms or automated phone systems.

This noise complaint solution can:

• Seamlessly integrate with existing public safety workflows, offering a range of capabilities to enhance operations.
• Efficiently classify and prioritize complaints so that urgent matters receive immediate attention.
• Connect seamlessly with dispatch systems, streamlining the process of assigning resources to incidents.
• Provide residents automated status updates, keeping them informed throughout the resolution process.
• Generate standardized reports, improving consistency, facilitating data analysis, and providing access to relevant information based on jurisdiction—which means that responders have the most pertinent data at their fingertips.
• Enhance the overall efficiency and effectiveness of public safety operations.

Implementation approach

When implementing generative AI solutions like our noise complaint solution, agencies should begin by clearly defining specific use cases that align with their operational needs. It’s crucial to maintain robust human review capabilities throughout the implementation process to provide accuracy and appropriate oversight. Privacy considerations must be thoroughly addressed, including data handling protocols and compliance with relevant regulations. Additionally, organizations need to invest time in preparing their staff for evolving roles, providing necessary training and support as responsibilities shift with the introduction of new technology. Through this responsible and methodical implementation approach, public safety agencies can achieve significant improvements in both community satisfaction and operational efficiency, ultimately enabling their personnel to focus more effectively on their core public safety mission.

Benefits to public safety

Early implementations of our generative AI-based noise complaint solution have delivered significant benefits, including:

• 40 percent reduction in dispatcher time spent on noise complaints
• Improved compliance with information gathering requirements
• 24/7 availability without additional staffing
• Better data for strategic resource deployment

Other applications

Beyond noise complaints, solutions using Amazon Bedrock Agents can help public safety agencies handle other issues such as:

• Reporting of traffic incidents without injuries
• Abandoned vehicle reporting
• Graffiti and vandalism reporting
• Lost property registration
• Public space maintenance issues

Each of these applications helps shift the burden of routine tasks from officers and staff to AI assistants—while maintaining human oversight where needed.

Conclusion

In this post, we discussed the value of using Amazon Bedrock Agents to create a generative-AI based noise complaint solution, which is capable of handling routine interactions as part of a comprehensive public safety system. To learn more about how to keep everyday tasks from overwhelming your department’s resources, schedule a free consultation with our public safety solutions team or request a demo of Amazon Bedrock Agents.