Guidance For Agentic Data Exploration On Aws

Guidance for Agentic Data Exploration on AWS

[!NOTE] Visit the [**Universal Knowledge Graph for Manufacturing**](https://github.com/aws-samples/sample-ukg-for-mfg) for the related project presented at Hannover Messe 2026.

Table of Contents

1. Overview
   • Cost
2. Prequisites
3. Notes
4. CloudShell Deployment
5. Local Deployment (Mac / Linux)
6. Deployment Validation
7. Running the Guidance
8. Evaluation and Testing
9. Tracing and Observability
10. Next Steps
11. Cleanup
12. Common issues, and debugging
13. Revisions
14. Authors

Overview

The Guidance for Agentic Data Exploration on AWS (Panoptic) is a Generative AI powered solution that leverages [Strands Agents SDK](https://strandsagents.com/) and [Amazon Bedrock](https://aws.amazon.com/bedrock/) to unify and analyze diverse data streams without traditional ETL barriers and data integration. This Guidance addresses the challenge of analyzing complex, interconnected data from multiple sources by providing a multi-agent system that can intelligently process, visualize, and extract insights from various data formats.

**Why did we build this Guidance?**

Traditional data analysis often requires extensive ETL processes and data integration efforts before meaningful insights can be extracted. This Guidance eliminates these barriers by providing specialized AI agents that can work collaboratively to process raw data, create graph representations, generate visualizations, and provide intelligent analysis without requiring upfront data transformation.

The system is comprised of specialized AI agents, each designed for specific use purposes:

The Guidance for Agentic Data Exploration on AWS (Panoptic AI Data Expl