Booking.com uses Amazon SageMaker and Amazon Bedrock to power personalized travel recommendations

Booking.com uses AWS to handle more than 150 petabytes of data and apply generative AI and fine-tuning to deliver relevant, personalized travel suggestions. The company collaborated with AWS Professional Services to build a scalable machine-learning framework that accelerates experimentation and innovation for destination and accommodation recommendations. The article presents Booking.com as using AWS to improve booking services globally across its travel marketplace.

Organization
Booking.com
Location
Netherlands
Published
May 2026

Reported outcomes

Strategic outcomes

New product / capabilityDelivered personalized travel recommendationsSpeed & agilityAccelerated experimentation and innovationCustomer experience & trustImproved booking services with generative AI

Primary read

Use case focus

Showing 3 of 4

  • 1Personalization
  • 2Recommendation Engine
  • 3Customer Experience
  • Deliver relevant, personalized destination and accommodation recommendations at global scale.
  • Use a scalable machine-learning and generative AI workflow across more than 150 petabytes of data.
  • Accelerate experimentation and innovation for booking services.
  • Booking.com uses Amazon SageMaker and Amazon Bedrock with AWS Professional Services to build and train machine-learning models for recommendations.
  • The AWS-based framework supports generative AI and fine-tuning language models for personalized travel suggestions.
  • The implementation is positioned as a scalable ML framework for Booking.com's global marketplace.
  • Improves booking services with generative AI.
  • Enables more relevant, personalized suggestions for travelers.
  • Accelerates experimentation and innovation at scale across a large marketplace.
Implementation partners1
Sources & evidence1
Groundedness: 4/5Type: Customer StoryPublished: May 13, 2026Publisher: AWSEvidence: PrimaryConfidence: High

AI-generated summary. Verify important details with the linked sources before relying on this case.

Explore related AI use cases

Was this useful?

Similar cases