Build AI-powered dashboard automation agents with NLP on Amazon Bedrock AgentCore

Artificial Intelligence Build AI-powered dashboard automation agents with NLP on Amazon Bedrock AgentCore Business analysts often wait days for dashboard modifications when responding to changing business requirements. Traditional processes involve submitting modification requests to IT teams, who interpret requirements, navigate API documentation, understand table schemas, and deploy changes. While this approach maintains proper oversight and quality control, it can result in multi-day turnaround times when rapid dashboard updates are needed. This solution combines the power of Amazon Bedrock AgentCore, Strands Agents, and Amazon Quick transforms to deliver a secure, scalable, and intelligent system for building and operating AI agents while transforming data into actionable business insights. Solution overview In this solution, we use a multi-agent architecture built with Amazon Bedrock AgentCore and the Strands framework. Amazon Bedrock AgentCore is an agentic platform for building, deploying, and operating effective agents securely at scale, no infrastructure management needed. It accelerates agents to production with intelligent memory and a gateway to enable secure, controlled access to tools and data. It runs agents with production-grade security and dynamic scaling and monitors performance and quality in production. Strands Agents is a code-first framework for building agents with integration to AWS services. The solution also uses Amazon Quick which delivers AI-powered BI capabilities, transforming your scattered data into strategic insights for everyone so you can make faster decisions and achieve better business outcomes. The architecture comprises three specialized agents working together. The Find Dashboard Agent performs discovery operations including searching dashboards and retrieving column metadata from dashboards and datasets. The Modify Dashboard Agent executes configuration changes by validating columns, updating table visuals, and creating new dashboard versions. The Orchestrator Agent routes user requests to the appropriate specialized agents based on intent classification. The Orchestrator Agent serves as the entry point for user interactions. When users submit natural language queries like “Add lastname to the testing dashboard”, Amazon Nova classifies requests as conversational or operational. Conversational queries receive direct responses using Nova’s large language model (LLM) capabilities. Operational requests are routed through the Strands framework to specialized agents, validates changes against available dataset columns, and executes modifications autonomously while maintaining security controls, audit trails, and preserving original dashboards for rollback purposes.The following diagram illustrates the solution architecture and workflow. The architecture includes the following components: - Amazon Bedrock AgentCore – Hosts the Strands Agent orchestrator and specialized sub-agents. - Amazon Nova – Provides natural language processing (NLP) and reasoning capabilities. - Amazon Quick – The target service for dashboard discovery and modification operations. - AgentCore Memory – Maintains conversation context and session state. - Amazon Bedrock AgentCore Observability – Logs agent decisions and traces API interactions. To implement the agentic AI solution for Quick self-service, complete the following high-level steps: - Build the agents (Find Dashboard Agent, Modify Dashboard Agent, and Orchestrator Agent). - Deploy the agents to Amazon Bedrock AgentCore. - Test the agent through the AWS Management Console. Prerequisites To implement this solution, you must have the following prerequisites: - An AWS account with permissions…