PRISM uses GenAI to make their extensive institutional knowledge more accessible for their members

Overview

The Cal Poly Digital Transformation Hub (DxHub), powered by Amazon Web Services (AWS), partnered with PRISM (Public Risk Innovation, Solutions, and Management) to solve a knowledge accessibility challenge. PRISM, established in 1979, is a Joint Powers Authority that provides risk management services to over 2,200 public entities across California, including counties, cities, educational organizations and special districts.

PRISM needed a better way to make their extensive institutional knowledge and resources more accessible and useful to their diverse membership base. The approach involved implementing generative AI technologies, to create a system capable of understanding complex queries, synthesize information from multiple sources within PRISM’s archives, and deliver tailored and citable responses to member inquiries.

The engagement between the Cal Poly DxHub and PRISM represents a significant step forward in PRISM’s digital transformation journey, reinforcing its position as a leader in innovative risk management solutions.

Problem

 

PRISM, while having an extensive knowledge repository, faced three main challenges in making their information accessible to members:

  1. Information Access: Members struggled to quickly find specific information within PRISM’s large resource library on their website, leading to inefficient searching through numerous documents.
  2. Multi-Format Content Challenges: PRISM’s knowledge base contained various formats (documents, videos, podcasts), making it difficult for members to extract insights across different media types.
  3. Content Digestibility: Members appreciate a way to get quick, concise summaries and specific answers without having to review entire documents or media files.

The DxHub needed to develop a solution that could:

  • Efficiently search and summarize content across multiple formats
  • Provide an intuitive user interface
  • Deliver concise answers with relevant source links
  • Route queries to appropriate departments when needed
  • Include proper disclaimers for sensitive information

These challenges highlighted the need for an intelligent, integrated solution that could not only search and summarize PRISM’s diverse content but also provide an intuitive interface for members to interact with this wealth of knowledge. The focus was to create an intelligent system that could make PRISM’s extensive knowledge base more accessible, actionable, and easily digestible for their members.

Innovation in Action

 

To address PRISM’s challenges, the Cal Poly Digital Transformation Hub (DxHub), powered by AWS, collaborated with PRISM to develop an innovative AI-driven solution. At its foundation, the system uses sophisticated content processing that can understand and integrate diverse formats – from traditional documents to audio files and video presentations – making PRISM’s entire knowledge base semantically searchable.

The heart of the innovation lies in its intelligent search and summarization capabilities. Members can now ask questions in plain language through a user-friendly chatbot interface and receive immediate, relevant responses with direct links to source materials. This dramatically reduces the time spent searching through extensive documentation while ensuring accuracy and reliability.

What makes this solution particularly powerful is its intelligent handling of complex situations. When queries require specialized expertise in areas like underwriting or actuarial analysis, the system automatically routes members to appropriate human experts. Additionally, it incorporates dynamic safeguards, generating context-specific disclaimers for sensitive information and providing relevant training recommendations when educational content matches a member’s query.

This thoughtful integration of advanced AI technology with human expertise creates a seamless experience that enhances PRISM’s ability to serve its members efficiently while maintaining appropriate oversight and guidance. The solution transforms what was once a time-consuming search process into an intuitive, responsive system that delivers precise, actionable information when members need it most.

Technical Solution

The innovative solution leverages a suite of AWS services to create a robust and scalable architecture that easily integrates via REST calls with PRISMs existing web technology. The processing of documents is achieved through leveraging assets on AWS S3 and AWS Lambda functions that continuously update and expand the repository as new content becomes available.

To process the diverse range of content, including text, audio, and video, the system employs an advanced chunking and embedding strategy. AWS Lambda functions work in conjunction with Amazon Titan Text Embeddings V2 to break down and semantically encode all ingested information. This approach allows for efficient retrieval and understanding of content across various formats.

For audio and video content, AWS Transcribe is utilized to convert spoken words into text, making this information as searchable and accessible as written documents. Additionally, Amazon Textract provides optical character recognition (OCR) capabilities, extracting text from slides and visual elements within video content, further enriching the knowledge base.

The heart of the system’s intelligence uses a combination of Amazon OpenSearch Service for efficient information retrieval and Amazon Bedrock, leveraging the Anthropic Claude Sonnet 3-5 model. As a result, the system can retrieve information from multiple sources, and generate accurate, context-aware responses.

To integrate with PRISM’s existing website, the solution utilizes API Gateway and AWS Lambda functions with Step Functions, to orchestrate the whole process. The entire system is designed with security, scalability, and performance in mind, making full use of AWS’s managed services to reduce operational overhead and ensure high availability.

Supporting Documentation

 

Source CodeAll of the code and assets developed during the course of creating the prototype.
Architecture DiagramA diagram that describes the technical components needed to implement the solution.

About the DxHub

The Cal Poly Digital Transformation Hub (DxHub) is a strategic relationship with Amazon Web Services (AWS) and is the world’s first cloud innovation center supported by AWS on a University campus. The primary goal of the DxHub is to provide real-world problem-solving experiences to students by immersing them in the application of proven innovation methods in combination with the latest technologies to solve important challenges in the public sector. The challenges being addressed cover a wide variety of topics including homelessness, evidence-based policing, digital literacy, virtual cybersecurity laboratories and many others. The DxHub leverages the deep subject matter expertise of government, education and non-profit organizations to clearly understand the customers affected by public sector challenges and develops solutions that meet the customer needs.