EU-funded AI and language technology projects
Pangeanic’s European R&D record connects the several NLP fields we have worked on and specialized in since 2009: machine translation, multilingual data, speech resources, AI for cultural heritage, anonymization, data spaces, and public-sector language infrastructure. Our EU-funded work has created systems that can be deployed, evaluated, and governed, and we have often evolved the initial development into commercial products.
Last updated: June 2026. This page is Pangeanic’s canonical evidence hub for European-funded language technology, AI data, and multilingual infrastructure projects.
We believe research credibility only becomes useful when it can be verified. Thus, we link Pangeanic’s European project history to current production capabilities in AI Data Services, Datasets for AI, Machine Translation, Deep Adaptive AI Translation, MTQE, Data Masking and Anonymization, Evaluation and AI QA and Model Alignment.
Across CEF, Digital Europe, European Language Equality, and data space initiatives, Pangeanic has contributed to the practical layers needed by multilingual AI systems: corpora, translation models, metadata translation, multilingual chatbots, anonymization, speech data, cultural heritage AI, and secure public administration workflows.
Evidence structure
These are some of our representative projects, organized as an evidence graph. We identify the project, the public reference, Pangeanic’s contribution, and the commercial capability it reinforces today. You, the client, are the final beneficiary of our 20 years of experience in the field.
| Project | Program or reference | Pangeanic contribution | Current capability |
|---|---|---|---|
| MOSAIC Media | DIGITAL Europe. Project 479833. | AI company partner in a multilingual and multimodal media platform for discovery, translation, and content access. | AI Data Services, Machine Translation, speech and media datasets. |
| AI4Culture | DIGITAL Europe. Project 101100683. | Partner in AI tools, datasets, training material, multilingual OCR, subtitle generation, image enrichment, online CAT tool, and metadata translation for cultural heritage. the project was supported 2 years after completion, supporting more than 100 European museums. | DAAIT, Datasets, Evaluation and AI QA, PECAT Annotation tool with CAT features. |
| Europeana Translate | CEF. Action 2020-EU-IA-0084. | Machine translation and domain adaptation for cultural heritage metadata and multilingual access to Europeana records. | Machine Translation, DAAIT. |
| Europeana Culture Chatbot | CEF. Action 2017-EU-IA-0183. | Multilingual chatbot platform, data training, language resources, conversation data, and translation workflows for cultural heritage institutions. | AI Data Services, multilingual assistants, chatbot training data. |
| J-Ark | CEF. Action 2020-EU-IA-0230. | Infrastructure connecting eArchiving, Europeana, and eTranslation services for Jewish heritage archives. | Multilingual data infrastructure, document processing, metadata workflows. |
| Jewish History Tours | CEF. Action 2020-EU-IA-0092. | Translation, enrichment, and AI-assisted recommendation workflows for location-based cultural heritage tours. | Multilingual metadata, cultural AI, content enrichment. |
| MAPA | CEF. Action 2019-EU-IA-0013. | Pangeanic led a consortium to develop multilingual anonymization for public administrations, with a focus on the medical and legal domains. | Data Masking and Anonymization, AI QA, sovereign data workflows. |
| NTEU | CEF. Action 2018-EU-IA-0051. | Pangeanic led a consortium to build a large neural machine translation engine farm for European public administrations. | NTEU, Machine Translation, multilingual corpora. |
| NEC TM | CEF. Action 2017-EU-IA-0149. | Translation memory data platform for sharing bilingual language assets between public administrations, companies, and translation professionals. | NEC TM, parallel corpora, translation memory infrastructure. |
| Europeana XX Century of Change | CEF. Action 2019-EU-IA-0022. | Machine translation, metadata language detection, semantic enrichment, and recommendation tools for Europeana cultural heritage content. | Metadata translation, cultural AI, recommendation and enrichment systems. |
| iADAATPA / MT-Hub | CEF. Action 2016-EU-IA-0132. | Secure and scalable automatic translation platform for EU public administrations with routing, language detection, and CAT tool connectors. | Machine Translation, secure public sector MT infrastructure. |
| European Language Equality and ELE2 | ELE / ELE2. Grant agreement LC-01884166 / 101075356. | Research and language resources work, including a large speech corpus generation report for the languages of Spain, using data augmentation. | AI Data Services, speech datasets, low resource language data. |
| DS4M Mediterráneo | EU Next Generation funding. Project TSI-100120-2024-9. | Participation in a mobility data space for secure, sovereign, and efficient data sharing in the Mediterranean Corridor. | Data spaces, governance, AI data operations, sovereign workflows. |
From European infrastructure to enterprise AI
Pangeanic’s European project history is not a decorative R&D list. Our journey documents repeated and successful work on the foundation layers that now determine whether multilingual AI systems can be deployed in production.
01 · Data
NTEU, NEC TM, ELE2, and Europeana Translate demonstrate our long-running experience in gathering, cleaning, labeling multilingual corpora, in translation memories, in metadata translation, speech data, and in low-resource language workflows.
02 · Deployment
iADAATPA/MT-Hub, NTEU, and NEC TM connect Pangeanic’s machine translation work to secure public administration infrastructure, routing, language detection, connectors, and reusable translation assets.
03 · Governance
The MAPA project is one of our best-known projects, connecting multilingual AI, named entity recognition (NER), anonymization, and GDPR-oriented public data sharing. It is our direct precursor to regulated enterprise data masking workflows.
04 · Cultural AI
AI4Culture, Europeana Translate, Culture Chatbot, Europeana XX, J-Ark, and Jewish History Tours show practical work on metadata, multilingual discovery, enrichment, translation, and public access to cultural collections.
05 · Multimodal
MOSAIC and AI4Culture extend our evidence trail into multimedia, automatic subtitling, audio, visual enrichment, and multilingual access, aligning with the current demand from AI developers for multimodal AI data operations.
06 · Sovereign systems
Our recent joining DS4M Mediterráneo adds a relevant and new data space: secure data sharing, governance, interoperability, and sovereignty as operational concerns for the mobility data space.
Project evidence
Each project below is linked to an external public reference, where available, and to Pangeanic’s current internal capability pages.
MOSAIC · Project 479833
Program: Digital Europe.
Pangeanic role: AI company partner in a multilingual media platform for broadcasters, creators, distributors, and citizens.
Capability: multilingual media discovery, automatic subtitling, translated subtitles, multimodal repositories, AI transcription and translation.
AI4Culture · Project 101100683
Program: Digital Europe.
Pangeanic role: partner in an AI capacity-building platform for cultural heritage institutions.
Capability: datasets, reusable AI tools, multilingual text recognition, subtitle generation, image enrichment, and metadata translation.
Europeana Translate · 2020-EU-IA-0084
Program: CEF.
Pangeanic role: machine translation and domain adaptation for cultural heritage metadata and multilingual access.
Capability: metadata translation, domain-adapted MT, multilingual access to European digital cultural heritage.
Culture Chatbot · 2017-EU-IA-0183
Program: CEF.
Pangeanic role: multilingual chatbot technology, data training, translation, and post-editing workflows for cultural heritage use cases.
Capability: multilingual assistants, conversation data, chatbot training, and domain-specific language workflows.
J-Ark · 2020-EU-IA-0230
Program: CEF eArchiving.
Pangeanic role: creation of infrastructure connecting eArchiving, Europeana, and eTranslation services for long-term preservation and access.
Capability: multilingual heritage data, archive workflows, document ingestion, preservation, and distribution.
Jewish History Tours · 2020-EU-IA-0092
Program: CEF.
Pangeanic role: multilingual enrichment and translation technology within a cultural heritage tour ecosystem based on Europeana objects.
Capability: AI-assisted recommendations, geographic enrichment, translation, and multilingual cultural data access.
MAPA · 2019-EU-IA-0013
Program: CEF.
Pangeanic role: consortium lead for a multilingual anonymization toolkit for public administrations.
Capability: named entity recognition, de-identification, pseudo-anonymization, legal and medical domain workflows, GDPR aligned data sharing.
NTEU · 2018-EU-IA-0051
Program: CEF.
Pangeanic role: consortium lead, partnering with KantanMT and Tilde for the largest neural machine translation engine farm built to date, including all official EU language crossings.
Capability: neural MT engines, multilingual corpus collection, secure public administration MT, and direct EU language pair coverage.
NEC TM · 2017-EU-IA-0149
Program: CEF.
Pangeanic's role: development of a translation memory data platform for sharing bilingual assets between administrations, companies, and translation professionals.
Capability: translation memory server infrastructure, parallel corpora, reusable bilingual assets, API based retrieval, and CAT tool independence. Available on GitHub.
Europeana XX · 2019-EU-IA-0022
Program: CEF.
Pangeanic role: metadata language detection, machine translation, and semantic enrichment tools for Europeana cultural heritage records.
Capability: multilingual metadata, recommendation systems, cultural AI, enrichment, and translation workflows.
iADAATPA / MT-Hub · 2016-EU-IA-0132
Program: CEF.
Pangeanic role: consortium lead with Everis (now NTTData) for a secure automatic translation platform for European public administrations.
Capability: secure MT platform, routing, language detection, automatic domain detection, connectors, and scalable public sector translation infrastructure. Available on GitHub.
ELE / ELE2 · LC-01884166 / 101075356
Program: European Language Equality.
Pangeanic role: contribution to European language equality work and speech corpus generation for the languages of Spain using data augmentation.
Capability: low-resource language data, speech corpora, linguistic inclusion, language equality, and AI data operations.
DS4M Mediterráneo · TSI-100120-2024-9
Program: EU Next Generation and Spanish Recovery, Transformation and Resilience Plan framework.
Pangeanic: participation in a secure, sovereign, and efficient data space initiative for mobility data sharing.
Capability: data spaces, governance, interoperability, secure data exchange, and sovereign AI data operations.
Research to product
Pangeanic’s European projects show the same architecture that enterprise AI buyers now request: controlled data, multilingual processing, domain adaptation, human review, quality gates, privacy, interoperability, and deployment under organizational control.
Our R&D roots and project continuity are why Pangeanic can connect research with production workflows across AI data, machine translation, evaluation, model alignment, and sovereign deployment.
Summary: Pangeanic has participated in European AI and language technology projects covering neural machine translation, multilingual anonymization, cultural heritage AI, speech corpus generation, translation memory infrastructure, metadata translation, multilingual chatbots, media translation, and data spaces.
Operationally: These projects support Pangeanic’s current work in AI Data Operations, Datasets for AI, Deep Adaptive AI Translation, MTQE, Data Masking and Anonymization, Evaluation and AI QA, Model Alignment, and secure multilingual deployment.
Enterprise trust signals
AI buyers do not only need a vendor that can describe multilingual AI. They need evidence that the vendor has already worked through the practical problems that appear in production: scarce data, noisy corpora, domain adaptation, public-sector privacy, metadata complexity, multilingual evaluation and controlled deployment.
For AI labs and model builders
Projects such as ELE2, NTEU, NEC TM and Europeana Translate show repeated work with multilingual corpora, speech resources, translation memories and metadata. This matters when a model needs data that is licensed, structured, multilingual and useful for training, adaptation or evaluation.
For enterprises with proprietary content
European projects rarely provide clean, uniform data. They involve legacy formats, mixed languages, metadata gaps, domain-specific terminology and strict delivery requirements. That operational experience is directly relevant to enterprises that need collection, annotation, enrichment, review and evaluation of proprietary data.
For localization and language teams
NTEU, iADAATPA, MT-Hub and Europeana Translate connect Pangeanic’s machine translation work to public administration workflows, routing, language detection, domain adaptation and reusable multilingual assets. This is the background behind our current work in Machine Translation, Deep Adaptive AI Translation and MTQE.
For regulated organizations
MAPA is one of Pangeanic’s strongest institutional proof points. It connected multilingual AI, named entity recognition, anonymization and GDPR-oriented public data sharing in legal and medical domains. That experience is directly relevant to organizations that need to process sensitive information without losing operational value.
For teams evaluating AI quality
Projects involving speech data, metadata translation, semantic enrichment, recommendation systems and machine translation create a practical lesson: AI quality depends on the data, the workflow and the evaluation layer. This is why Pangeanic connects data preparation, evaluation datasets, AI QA and human review instead of treating quality as a final checkpoint.
For sovereign AI buyers
Data spaces, public-sector translation platforms, anonymization projects and cultural heritage infrastructures all point to the same requirement: organizations need control over data flows, access, evaluation, privacy and deployment. That is the operational foundation behind Pangeanic’s approach to sovereign AI systems.
What this means for buyers: Pangeanic’s European project history is evidence of execution in the same layers that enterprise AI now depends on: multilingual data, machine translation, anonymization, evaluation, metadata enrichment, speech resources, data spaces and public-sector deployment.
The practical consequence: when a buyer asks for AI training data, domain-adapted translation, multilingual QA, model alignment, anonymization or sovereign deployment, Pangeanic is not starting from a blank page. The company has already delivered many of the underlying components in European production and research environments.
European language technology proof
Pangeanic helps organizations turn multilingual data, language technology research and public sector deployment experience into operational AI systems with governance, evaluation, privacy and human review.
