altt

Service Referral Agent: Brightpoint

Overview

The Arizona State University Artificial Intelligence Cloud Innovation Center, powered by Amazon Web Services (AWS), collaborated with Brightpoint, a community-based, non-profit organization that provides prevention-based services centered around the family to prevent moments of instability from becoming life-altering crises. Brightpoint works with families and their children and youth in Illinois, providing early education and family support, mental health and wellness, and child welfare services. Together, we developed the Brightpoint Service Referral Agent—a responsive web-based and SMS-enabled chatbot designed to give every family in Brightpoint’s Doula and Home Visitor Programs expanded, 24/7 access to critical support services and community resources they need to thrive.

Problem

Families struggle with fragmented referral processes, lack of knowledge of services available to them and overwhelming paperwork. While the Brightpoint staff serve as a trusted resource, they lack real-time visibility into client needs, service demand, and fluctuating availability of community resources. 

  • Fragmented Channels: Families navigate multiple systems leading to missed referrals and follow-up challenges
  • High Staff Workload: Home visitors spend excessive time manually researching and delivering referrals, reducing direct client support
  • Lack of Insights: Brightpoint currently cannot track which services are most requested or whether community needs are fully met, and how clients use the services

Student Spotlight

The AI CIC is powered by ASU Student Workers. The following students were assigned to this project to develop this open-source solution in partnership with the AWS and ASU mentor team.

Approach

The ASU CIC team delivered a multi-channel, serverless solution that automates referrals, captures feedback, and provides admin insights. Brightpoint Referral Agent is a responsive web-based chatbot designed to assist families in Brightpoint’s Doula and Home Visitor Programs. It helps users navigate essential support services like mental health, child welfare, and youth programs through a simple conversational interface. The application also provides Brightpoint’s internal staff (admins) with tools for monitoring and managing referrals via an admin portal.

  • Web & SMS Chat: React.js portal hosted on AWS Amplify; SMS integration via Amazon Pinpoint and Lambda
  • Authentication & History: Amazon Cognito for secure sign-in, user profiles, and saved conversation transcripts
  • Referral Orchestration: Bedrock-based agent queries DynamoDB for known referrals; falls back to Perplexity LLM for unmatched queries
  • Multilingual Support: Automatic translation pipeline for English, Spanish, and Polish queries
  • Admin Dashboard: Role-based CRUD operations on referrals, feedback review, and trend analytics

Industry

By automating referral delivery and feedback capture, the Brightpoint Service Referral Agent empowers families to access critical resources instantly, while freeing home visitors to focus on meaningful, in-person support. This data-driven platform not only improves response accuracy but also uncovers actionable insights, enabling Brightpoint to optimize community resource allocation and better meet evolving client needs.

"With the support of the AWS CIC team, we developed the Service Referral Agent to better serve families in our doula and home visiting programs. Families will no longer have to hold onto paper forms or wait for a worker to respond — they will receive immediate guidance to the resources they need. This means faster support during critical moments, while also improving data accuracy and strengthening our ability to track and report on the impact of our services.” 

Tiffanny Powell, Senior Director, Brightpoint

Application

This open-source framework can be adopted by other social-service agencies, healthcare providers, and community organizations to automate personalized referrals, extend support via web and SMS, and measure impact through real-time analytics.

Supporting Artifacts

 

 

 

Next Steps

The Brightpoint team is taking a methodical approach to building, testing, deploying, and maintenance of this solution. Here is its focused approach: 

  • Refine the Chatbot
       Adjust conversation flows based on feedback.
       Make sure all resources are up-to-date and easy to understand.
  • Test with Families
       Try it out with a small group of participants.
       Gather feedback to make sure it’s easy to use and helpful. 
  • Prepare Staff
       Train home visiting and doula staff on how it works.
       Create simple guides and materials to support staff and families.
  • Set Up Ongoing Support
       Assign a team to keep the chatbot updated and fix any issues.
       Regularly check that it’s working well and improving over time.
  • Measure Impact
       Track how families are using it.
       Use the data to understand how it’s helping and where to improve.

About the ASU CIC

The ASU Artificial Intelligence Cloud Innovation Center (AI CIC), powered by AWS is a no-cost design thinking and rapid prototyping shop dedicated to bridging the digital divide and driving innovation in the nonprofit, healthcare, education, and government sectors.

Our expert team harnesses Amazon’s pioneering approach to dive deep into high-priority pain points, meticulously define challenges, and craft strategic solutions. We collaborate with AWS solutions architects and talented student workers to develop tailored prototypes showcasing how advanced technology can tackle a wide range of operational and mission-related challenges. 

Discover how we use technology to drive innovation. Visit our website at ASU AI CIC or contact us directly at [email protected].

Photos

5
6
4
3
1
2