Home icon

How Bark.com and AWS collaborated to build a scalable video generation solution

Machine Learning Blog



This article describes how Bark.com and AWS built a scalable AI-powered video generation solution using Amazon Bedrock and SageMaker to reduce marketing content production from weeks to hours.

  • Reduced production time from weeks to hours for personalized social media ad campaigns
  • Uses Amazon Bedrock (Claude Sonnet 3.7) for text generation and creative ideation
  • Implements Wan 2.1 text-to-video model on p4de.24xlarge SageMaker instances with 8-GPU tensor parallelism
  • Maintains visual consistency through semantic specification and reference image propagation
  • AI-generated ads scored 25% higher on originality and matched professional production quality
  • Generates 15-30 second ads in 12-15 minutes including quality checks and regeneration loops
  • Human-in-the-loop review at creative brief and final approval stages ensures brand alignment

The solution demonstrates a replicable architecture pattern for AI-assisted content production combining hierarchical prompt planning with reference image propagation to maintain consistency across multi-scene video generation.



Go to article

The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.

Related articles

Mar 16
2026
How Sony Pictures Networks India Delivered Personalized Ad Insertion at Scale on AWS
Mar 25
2026
Unlocking video insights at scale with Amazon Bedrock multimodal models
Feb 25
2026
How AWS Built a Live AI-Powered Vertical Video Capability for Fox Sports with AWS Elemental Inference
Mar 19
2026
Use RAG for video generation using Amazon Bedrock and Amazon Nova Reel

The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.