Extending Public Sector Intelligence with Agentforce and AWS

Serdar HocamAuthor & Editor

Amazon Web Services and Salesforce have introduced a new integration that turns unstructured data into actionable insights.

◉ 2 views
Extending public sector intelligence with Agentforce and AWS | Amazon Web Services

Amazon Web Services and Salesforce have announced an integration that transforms unstructured data—such as body camera footage, surveillance videos, and scanned documents—into structured, actionable insights with the help of Amazon Bedrock Data Automation, Model Context Protocol, and Salesforce Agentforce.

Data Processing Challenges in the Public Sector

Public sector organizations process massive volumes of unstructured evidence data, such as body camera footage, surveillance videos, and scanned documents, from which insights must be extracted before action can be taken.

The Importance of Storage and Automation

While durable and cost-effective storage is provided through the Amazon S3 integration, storage is only half the battle, and manual reviews take too much time without automation.

Access via the Salesforce Console

Through this integration, Agentforce users can search processed data stored on AWS, uncover key insights from unstructured data, and perform advanced operations without leaving the Salesforce console.

Workflows That Turn Raw Evidence into Insights

Two main workflows work together to transform raw evidence into actionable investigative insights; the first workflow uploads unstructured media files to Amazon S3.

Event-Driven Architecture on AWS

Once data arrives in Amazon S3, an event-driven architecture asynchronously processes multi-modal data using Amazon Bedrock Data Automation and organizes the results.

Communication Between Agentforce and AWS

On the Salesforce side, chats conducted via Agentforce trigger a structured action that calls AWS via the Model Context Protocol, providing responses in natural language.