rabbit: migrating to Amazon EKS and using Amazon Bedrock (Claude) to power a generative AI device assistant
In 2024, technology startup rabbit launched the r1 handheld device, which uses voice, image, and text input to help users interact with the world around them. After a CES demand surge and preorder growth, rabbit needed scalable, secure infrastructure to support more than 100,000 devices, improve latency and rate quotas, expand geographic support, and reduce AI service costs. rabbit migrated from another cloud to a containerized infrastructure on Amazon Elastic Kubernetes Service (Amazon EKS) and optimized its AI stack on Amazon Bedrock with Anthropic Claude. With partner Caylent, rabbit also automated infrastructure management, added security controls, and prepared for over-the-air updates and future multicluster scaling.
- Organization
- rabbit
- Industry
- Consumer & Food
- Location
- United States
- Published
- May 2026
Reported outcomes
Strategic outcomes
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- rabbit
- Provider
- AWS
- Maturity
- Unknown
- Linked source
- AWS Customer Stories
No explicit deployment-stage evidence found.
Primary read
Use case focus
Showing 3 of 4
- 1AI-powered device assistant
- 2Cloud migration
- 3Infrastructure modernization
- After CES demand surged, rabbit needed to scale its large action model quickly and securely.
- The company needed to support far more devices than initially planned, while improving latency, geographic support, and cost efficiency.
- rabbit also needed to maintain strong security and reliability during rapid growth.
- rabbit migrated its infrastructure to Amazon Elastic Kubernetes Service (Amazon EKS) in about two months.
- The company adopted Amazon Bedrock with Anthropic Claude to power and optimize generative AI functionality for the r1 device.
- With Caylent, rabbit automated infrastructure management, implemented security controls, and began preparing a multicluster, cell-based architecture on AWS.
- The migration supported scaling from a planned 3,000 users to more than 100,000 r1 units.
- rabbit improved system performance and reliability and accelerated voice-operation response times.
- Using Amazon Bedrock helped reduce AI service costs and text-to-speech latency.
- The AWS setup enabled secure over-the-air updates and ongoing fleet expansion.
- The company met its launch deadline despite the hypergrowth surge.
Architecture
rabbit migrated from a different cloud provider to a containerized AWS environment centered on Amazon Elastic Kubernetes Service (Amazon EKS). The AI assistant stack was optimized with Amazon Bedrock and Anthropic Claude. Caylent helped orchestrate the migration, automate infrastructure management, implement security controls such as service control policies and AWS WAF, and prepare the environment for secure over-the-air updates and future multicluster, cell-based scaling.
Implementation partners1
Sources & evidence1
- Customer explicitly identified
- Primary source available
- Technical implementation details available
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Was this useful?
Community
Comments
Loading comments...