Agentic conversational video intelligence built on AWS
AWS shows how to build a multi-service agent that understands video in real time — without custom orchestration.
Why it matters
This is a vendor how-to demonstrating agent architecture in practice: a single agent coordinates three AWS services (Bedrock, Rekognition, Transcribe) to answer natural-language questions about video. Practitioners building agents on AWS will learn the pattern; it signals where AWS sees agent adoption heading.
The key facts
9 to knowStrands Agents SDK used for orchestration
Single agent coordinates multiple AWS services: Amazon Bedrock, Amazon Rekognition, Amazon Transcribe
Agent makes runtime decisions about which service to call
Use case: natural language queries on video returning answers in seconds
Published as AWS blog tutorial (vendor how-to)
Agentic architecture orchestrating multiple AWS services at runtime
Services: Amazon Bedrock, Amazon Rekognition, Amazon Transcribe
Use case: natural language Q&A over video
Published as AWS blog tutorial/how-to
The story so far
Earlier coverage of this storyline
Go to the source
AWS Machine Learning Blogaws.amazon.com
Publisher excerpt: Learn how to build a conversational video intelligence solution on AWS using an agentic architecture. A single Strands Agents SDK agent orchestrates Amazon Bedrock, Amazon Rekognition, and Amazon Transcribe at runtime, deciding which service to call so you can ask natural language questions about…