What challenges did Bundesliga face before AI implementation?

Editors relied on a time-consuming manual workflow to translate and localize match content into multiple languages. The league needed to increase language coverage and speed up distribution for a global fan base.

What was the impact of the AI solution?

Video localization processing time was reduced by 75%. Localization cost per match per language dropped from about $9.5 to $0.05, a reduction of more than 99%. Large language model inference cost for Content OS was reduced by 3.5x. Editors can spend less time on manual translation and more time creating richer content.

Source

Proof of conceptEvidence: Medium65/100

The Bundesliga transforms fan experiences with AI using Amazon Nova

Use case typeMultilingual communicationUpdated Jun 13, 2026

Bundesliga, operated by DFL Digital Sports in Germany, produces live and post-match content for more than 1 billion fans globally. The organization needed to localize match content into many languages faster and at lower cost than a manual editorial workflow could support.

Organization: Bundesliga
Industry: Other
Location: Germany
Published: April 2026

Reported outcomes

−75%

timeTime & speed

−99%cost3.5xcost

Strategic outcomes

Customer experience & trustImproved multilingual fan content deliveryNew product / capabilityBuilt real-time multilingual match insightsNew product / capabilityCreated automated content localization workflowNew product / capabilityAutomated archive metadata generation

Catalog median for time & speed deployments: −50% across 312 reported metrics. Compare benchmarks →

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Time: 75% decrease

AWS Customer StoriesApr 29, 2026Customer storyInferred claimMedium evidence strength

Video localization processing time was reduced by 75%.

Normalized claim

Cost: 99% decrease

AWS Customer StoriesApr 29, 2026Customer storyInferred claimMedium evidence strength

Localization cost per match per language dropped from about $9.5 to $0.05, a reduction of more than 99%.

Normalized claim

Cost: 3.5 x decrease

AWS Customer StoriesApr 29, 2026Customer storyInferred claimMedium evidence strength

Large language model inference cost for Content OS was reduced by 3.5x.

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Bundesliga, DFL Digital Sports
Provider: AWS
Maturity: PoC
Linked source: AWS Customer Stories

Bundesliga evaluated foundation models and ran a proof of concept before moving to production-ready tools

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 5

1Content localization
2Multilingual content generation
3Speech/transcription

Editors relied on a time-consuming manual workflow to translate and localize match content into multiple languages.
The league needed to increase language coverage and speed up distribution for a global fan base.

Bundesliga evaluated foundation models and ran a proof of concept before moving to production-ready tools.
The organization used Amazon Nova foundation models through Amazon Bedrock to build three production use cases: AI Live Ticker for real-time multilingual match insights, Content OS for transcription, translation, voice-over, subtitling, and distribution, and Intelligent Generation of Metadata for searching more than 50 years of archive content.
Human editors still review content before publication, but AI handles the majority of localization and metadata work.

Technologies

Amazon Nova Amazon Bedrock

Video localization processing time was reduced by 75%.
Localization cost per match per language dropped from about $9.5 to $0.05, a reduction of more than 99%.
Large language model inference cost for Content OS was reduced by 3.5x.
Editors can spend less time on manual translation and more time creating richer content.

Sources & evidence1

Evidence: Medium65/100Evidence strength