
Building AI with Moroccan heritage
Join our community to help build language datasets and improve AI models for Darija.
Review and rate the accuracy of Darija translations for various text data.
Translate provided text data into Darija to help build our language dataset.
Developing AI Models with Moroccan Identity
ATLASIA is an open-source initiative committed to developing AI models that resonate with Moroccan values, identity, and culture. Our user-friendly platform facilitates the collection and storage of data, ensuring it is accessible to the public to foster AI advancements in Morocco.
Our Teams
Bringing together diverse expertise to build AI solutions that reflect Moroccan culture and values.
AtlasIA emerged from the shared vision of four individuals passionate about AI projects relevant to the Moroccan culture. Their collaborative efforts led to the establishment of AtlasIA as an open-source project dedicated to developing culturally attuned AI models.
The WebDev Team is responsible for AtlasIA's digital presence, creating a sleek, user-friendly website and an innovative annotation platform. They consistently deliver high-quality digital solutions that enhance the user experience.
The Data Team focuses on the technical side of AI development, including data collection and curation, AI model training, and benchmark design. Their work ensures that the AI systems are culturally relevant to Morocco.
The Comm Team manages AtlasIA's public image, designing the visual identity, managing social media, and handling public relations. They effectively communicate AtlasIA's goals and achievements to the community and interested parties.
Join Our Community
Be part of the movement to create AI solutions that understand and respect Moroccan culture and values. Contribute your expertise or support our open-source initiatives.
Our Journey
Our strategic timeline to build and deploy Darija language models
- 1Build high-quality darija datasets
- 2Build darija benchmarks
- 1Build high-quality darija datasets
- 2Build darija benchmarks
- 1Machine translation models
- 2Other surprises
- 1Machine translation models
- 2Other surprises
- 1Model Deployment
- 2Continuous Improvement
- 1Model Deployment
- 2Continuous Improvement