What challenges did Strava face before AI implementation?

Help users make sense of complex workout data at scale. Generate personalized insights in Strava's brand voice. Maintain safety and quality for global user-generated content.

What was the impact of the AI solution?

The feature scales to 14 million tokens per minute at peak throughput. More than 80% of users who provided feedback said the insights were very helpful or helpful. Strava says Amazon Bedrock helps accelerate response times and avoid the need to self-host the model.

Rank#482of 3,107

Source

Strava safely transforms the athlete experience using Amazon Bedrock

Use case typeShopping recommendationsUpdated Jun 13, 2026

Strava delivers personalized insights through Athlete Intelligence powered by Claude in Amazon Bedrock. The feature helps users understand complex workout data such as pace, power, distance, and segments in a friendly, encouraging brand voice. Strava worked with AWS generative AI experts to select Claude Haiku and used Amazon Bedrock Guardrails to improve content safety and quality.

Organization: Strava
Industry: Other
Location: United States
Published: May 2026

Reported outcomes

+80%

quantified impactAdoption & scale

Strategic outcomes

New product / capabilityLaunched personalized athlete insights featureCustomer experience & trustImproved workout data understandingRisk & complianceImproved content safety and qualityScale & capacityScaled to high peak throughput

Catalog median for adoption & scale deployments: +85.5% across 46 reported metrics. Compare benchmarks →

Primary read

Use case focus

Showing 3 of 3

1Personalized recommendations
2Content moderation
3Customer engagement

Help users make sense of complex workout data at scale.
Generate personalized insights in Strava's brand voice.
Maintain safety and quality for global user-generated content.

Strava built Athlete Intelligence using Anthropic Claude Haiku in Amazon Bedrock to analyze activity stats and athlete-provided descriptions.
The system generates personalized, context-aware insights within seconds after an activity is completed.
Strava implemented Amazon Bedrock Guardrails to detect factual errors, filter harmful content, and mitigate undesirable topics.
Strava collaborated with the AWS Generative AI Innovation Center during model selection and launch.

Technologies

Amazon Bedrock Amazon Bedrock Guardrails Claude Haiku

The feature scales to 14 million tokens per minute at peak throughput.
More than 80% of users who provided feedback said the insights were very helpful or helpful.
Strava says Amazon Bedrock helps accelerate response times and avoid the need to self-host the model.

Architecture

Users complete an activity in the Strava app or upload from a connected device. Athlete Intelligence then analyzes the activity stats together with the athlete's activity history and athlete-provided descriptions. Claude Haiku in Amazon Bedrock generates personalized insights, and Amazon Bedrock Guardrails helps keep outputs safe and on-brand.

Implementation partners2

Anthropic AWS Generative AI Innovation Center

Sources & evidence1

Groundedness: 5/5Type: Customer StoryPublished: May 13, 2026Publisher: AWSEvidence: PrimaryConfidence: High

Primary source

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

Explore related AI use cases

Customer evidence