DataLab Group is a center of expertise dedicated to designing innovative and industrial-grade data and AI solutions. Its teams develop a wide range of AI use cases across multiple domains, relying on diverse types of data.
Since 2022, DataLab Group has been engaged in a certification and labeling approach to demonstrate that AI solutions can be both innovative and trustworthy. A core principle guides this work: to control AI systems, it is essential to fully understand and govern the data that powers them.
Challenges
AI initiatives within DataLab Group involve complex data landscapes and varied project requirements, creating several challenges:
- Centralizing documentation across numerous AI projects
- Managing multiple data types, including structured data, text, documents, and images
- Documenting data sources, processing steps, and indicators consistently
- Supporting certification and labeling processes for AI initiatives
- Scaling documentation practices across teams and projects
The teams needed a single solution capable of structuring data knowledge end-to-end, from raw data to AI use cases.
Why DataGalaxy
DataGalaxy was selected based on three key criteria:
- Ability to support a wide variety of AI use cases and data types
- Adaptability to different project contexts
- Simple and intuitive user experience, combined with responsive and available support teams
This flexibility made it possible to document AI projects in a consistent way, regardless of the underlying data.
“I’d recommend DataGalaxy for three main reasons: Ease of use, adaptability to different use cases, and efficient, available support teams. ”
DataGalaxy implementation
DataLab Group started with a pilot phase, retro-documenting three AI use cases as part of its certification process. Following this initial step, the approach was progressively extended to more projects.
The teams structured their documentation using templates within DataGalaxy. Three templates were created to reflect the specificities of structured data, text, and image data, while maintaining a shared common foundation.
Today, DataGalaxy is being generalized across AI projects, including those following formal certification and labeling processes.
Outcomes
With DataGalaxy, DataLab Group is able to:
- Centralize and document data sources, including usage licenses
- Describe how indicators are built from raw data, sometimes combining multiple data sources
- Structure complex data environments involving hundreds of raw data elements and indicators
- Support internal initiatives such as the distribution of open data across the organization
- Apply consistent documentation practices across AI projects
Key takeaway
Trusted and responsible AI relies on clear, structured data knowledge.
By centralizing and standardizing data documentation, DataGalaxy helps DataLab Group build AI solutions that are transparent, scalable, and aligned with trust and certification objectives.
