# @awseventschannel on YouTube

- **Type:** Video
- **Original URL:** https://youtube.com/watch?v=LnCXoIr0Mw8
- **Gondola URL:** https://gondola.cc/posts/65506628-awseventschannel-youtube
- **Thumbnail:** https://img.gondola.cc/tr:w-,h-,fo-auto/postThumbnails/d83cdb5e43.jpg
- **Posted:** 2026-05-15T05:54:55.000+00:00
- **Account Owner:** AWS Events (@awseventschannel) — https://gondola.cc/awseventschannel

## Caption

🕵️ Your AI agent shouldn't need a decision tree to decide how to investigate. 

The answers to complex questions often don’t live in a single data source. They span across multiple locations, domains, and vary in structure – tabular, connected, and unstructured, just to name a few.  

Agentic GraphRAG gives your agent a toolkit and lets it choose which source(s) would be best:

🔹 A Neptune MCP server for open-ended knowledge graph exploration 
🔹 Custom domain tools for defined, repeatable, and high-confidence subgraph extraction 
🔹 Auto-generated domain tools to query over unstructured documents  

We demo a fraud detection use case with #AmazonNeptune & #AmazonBedrock, but the architecture works for any domain where your #agenticAI needs to follow leads across structured & unstructured data. 

Stop orchestrating & start letting your agent investigate.

## Stats

- **Views:** 2,242
- **Likes:** 72
- **Shares:** 0
- **Comments:** 2

## Tags

amazonbedrock, agenticai, amazonneptune

---
Copyright (c) Gondola