01
Language fragmentation
European content must cross linguistic borders without depending on English as the only access route.
European media AI infrastructure
MOSAIC Media: multilingual AI infrastructure for European audiovisual content
MOSAIC connects broadcasters, media archives and technology providers through a multilingual and multimodal platform for content discovery, transcription, translation, subtitling, dubbing, summarization and metadata enrichment.
Digital Europe European Union funded media innovation project
LC-03303874 European Union Grant Agreement
5 broadcasters Public and private media organizations across Europe
Multimodal AI Audio, video, text, speech, subtitles and metadata
Project 479833 Pangeanic European project reference
The European media challenge
Valuable audiovisual content remains fragmented by language, format and infrastructure
European broadcasters hold extensive archives of news, documentaries, cultural programming and public interest content. Much of that material remains difficult to discover or reuse across national and linguistic borders.
Separate repositories, inconsistent metadata, limited subtitle coverage and uneven language technology support create practical barriers for broadcasters, creators, distributors and citizens.
MOSAIC addresses this fragmentation through a federated platform that combines media infrastructure with multilingual AI services and controlled content sharing.
02
Disconnected archives
Broadcaster repositories use different systems, metadata models and rights management practices.
03
High processing costs
Transcription, translation, subtitling and enrichment remain expensive when handled as isolated tasks.
04
Low resource languages
Smaller European languages often receive less technological support despite their cultural relevance.
Pangeanic in MOSAIC
From project requirements to an operational linguistic services layer
Pangeanic contributed at two connected levels: defining the foundations for a viable European media ecosystem and supplying the multilingual processing infrastructure required to operate it.
WP2 · Requirements and strategy
Functional, regulatory and commercial framework
Pangeanic led work on the requirements that shaped the MOSAIC ecosystem, connecting technical architecture with broadcaster needs, European regulatory obligations and future commercial sustainability.
- Functional requirements for broadcasters and media users
- Technical architecture and interoperability requirements
- Legal and regulatory considerations
- Ethical and trustworthy AI principles
- Performance metrics and impact indicators
- Commercial exploitation and sustainability criteria
WP3 · Integration and prototyping
Unified linguistic and multimedia API
Pangeanic deployed a decoupled services layer that allows MOSAIC nodes to orchestrate multilingual and multimedia tasks programmatically through secure endpoints.
- Speech to text for audio and video
- Multilingual translation of text and subtitles
- Text to speech and synthetic voice generation
- Automatic subtitling
- Summarization and editorial metadata generation
- Automatic dubbing between languages
Multimedia services
One API for multilingual audiovisual processing
Each capability is exposed as a service that can be combined into workflows according to the broadcaster, content type, target language and distribution requirement.
STT
Speech transcription
Convert audio and video speech into structured text for search, indexing, accessibility and downstream processing.
MT
Multilingual translation
Translate transcripts, subtitles, descriptions and editorial content across European languages.
SUB
Automatic subtitling
Produce timed multilingual subtitles that can support publication, accessibility and international reuse.
TTS
Synthetic voice
Generate speech from translated or adapted text for voice over, accessibility and multilingual delivery.
NLG
Summaries and metadata
Create summaries, descriptions, tags and editorial metadata that improve discovery and archive navigation.
DUB
Automatic dubbing
Combine transcription, translation and voice generation into controlled multilingual dubbing workflows.
Service oriented architecture
Modular workflows that connect broadcaster nodes with specialized AI services
Fincons coordinates the project and integrates the platform. Pangeanic and other technology partners provide specialized AI capabilities through secure and scalable interfaces.
Media source Audio, video, news, documentaries and archive content
MOSAIC node Dedicated broadcaster instance with controlled configuration
API gateway Secure orchestration between platform and AI tools
Pangeanic services Speech, translation, subtitles, summaries, voice and dubbing
Enriched content Searchable and reusable multilingual audiovisual assets
Each broadcaster operates a dedicated MOSAIC node. Content can remain under local control and may be shared with other nodes when the rights holder explicitly marks the asset as shareable.
European validation network
Developed with broadcasters, technology companies, researchers and media specialists
MOSAIC combines real broadcaster requirements with technical integration, multilingual AI and academic expertise across several European media environments.
Broadcasters
Technology, integration, research and market expertise
Research to production
What MOSAIC proves for media and enterprise buyers
MOSAIC demonstrates Pangeanic’s ability to connect multilingual AI services with real audiovisual infrastructure, organizational requirements, governance constraints and production workflows.
01
For broadcasters and media archives
Process, translate, enrich and expose audiovisual collections through modular workflows that respect repository ownership and content rights.
02
For streaming and content platforms
Add multilingual transcription, subtitles, synthetic voice, summaries and metadata through orchestrated API services.
03
For AI developers
Build or evaluate multimodal processing pipelines using speech, translation, media metadata and human quality controls.
04
For European institutions
Support multilingual public access, cultural diversity and controlled digital infrastructure across major and low resource European languages.
Public evidence
Official project information and partner references
The sources below document the project, consortium, Pangeanic contribution and broadcaster validation.
Official project
MOSAIC project website
Project objectives, consortium, technology, broadcasters and European Union funding reference.
View source →Pangeanic announcement
European TV and video portal
Initial description of the project, platform concept and Pangeanic’s role as a technical partner.
View source →Technical contribution
Pangeanic multimedia services layer
Account of the requirements work and multilingual processing services deployed by Pangeanic.
View source →Project coordinator
Fincons Group reference
Coordinator perspective on platform integration, consortium structure and technology partners.
View source →Broadcaster validation
3Cat live demonstration
Broadcaster evidence of platform demonstrations and practical multilingual audiovisual use.
View source →Pangeanic project hub
European AI and language technology projects
MOSAIC within Pangeanic’s wider record in multilingual data, media AI and European infrastructure.
Explore project history →External coverage
MOSAIC in broadcaster, institutional and industry sources
MOSAIC has been documented by participating broadcasters, research institutions, technology partners and audiovisual industry publications. These sources provide external validation of the consortium, live demonstrations and multilingual media objectives.
01
Broadcaster coverage
Project information and demonstrations published by participating European broadcasters.
3Cat First live demonstration of the MOSAIC audiovisual platform ↗ Sky Italia MOSAIC project and multilingual European media access ↗ Deutsche Welle A European video streaming platform across languages ↗ VRT MOSAIC within VRT’s European media innovation program ↗ RTV Slovenija RTV Slovenija participation in the MOSAIC project ↗
02
Consortium and research partners
References from the project coordinator, research institutions and technology partners.
Fincons Group European MOSAIC project launch and consortium structure ↗ Universitat Autònoma de Barcelona European project to increase linguistic diversity in audiovisual content ↗ CARSA Building a unified European television and video platform ↗ Official MOSAIC website Project objectives, partners and technology architecture ↗
03
Industry and media coverage
Audiovisual and technology media reporting on MOSAIC demonstrations and multilingual AI.
04
Institutional and sector validation
European language technology and data infrastructure material that places Pangeanic’s work in a broader multilingual AI context.
Multilingual media AI
Build accessible audiovisual workflows across languages
Pangeanic supports broadcasters, content platforms, institutions and AI developers with multilingual speech processing, translation, subtitles, metadata, evaluation and multimodal data operations.