Data to be mastered
Digital transformation continues and accelerates the evolution of the information landscape:
- Data volumes are increasing, flows and formats are increasingly heterogeneous, storage is diversifying and hybrid under the influence of cloud technologies.
- Data analysis is becoming more democratic by opening up to new profiles, particularly business users. Despite a lack of knowledge of traditional tools, data collection, enrichment and transformation are accelerating.
At the same time, management constraints are increasing, whether it is:
- Internally with increased needs for data quality and traceability.
- Externally with stronger regulatory constraints: on types of data (personal data for example with the DGPS) on specific sectors (BCBS 239)...
Technologies and organizations are continuing their transformation to address this evolution, but a challenge remains: data knowledge. Without it, there can be no recovery or compliance.
Major challenges:
- Lack of knowledge: In the absence of a centralized and ergonomic repository, data experts spend most of their time searching, understanding and preparing relevant data.
- Misunderstanding: teams try to collaborate, spreadsheets are created, but knowledge and collaboration are not shared.
- Scalability: data evolves (very) quickly. Too often, macro mapping of IT systems reflects a frozen top-down view which struggles to adapt to data agility.
DataGalaxy, agile data knowledge management platform
We found that the various initiatives to address this lack of knowledge are struggling to materialize. The implementation is indeed complex, it is a question of reconciling several problems which do not have the same temporalities or the same objectives: good management (responsibility, definitions, policies), daily actions, agility, transversality of data, diversity of actors...
We developed DataGalaxy with the ambition of operationalizing data management in order to initiate governance.
The objective of the platform is to build a 360° view of the data by answering 4 questions:
- What? The data dictionary reconciles the technical and functional definitions of the data in order to know the meaning of each business object and in particular its conditions of use.
- How? The processing repository details the flows and transformations carried out between different databases in order to control data flows and inter-application impacts.
- Where? The database dictionary describes the databases, tables, and fields and also allows them to be modeled in order to have an exhaustive view of the information heritage.
- For what? For whom? The usage repository catalogs all data uses: reporting, screens, extractions to find out why the data is being used and by whom. For more details, see this article.
Two elements are essential to build this vision and ensure its sustainability:
- Collaborate: Our solution aims to recover the inputs of each stakeholder, in a crowdsourcing approach, in order to establish a vision as exhaustive and shared as possible.
- Iterate and scale: Our solution is built to be able to map perimeters that are easy to understand, which proves the added value at each step and keeps teams motivated by the frequent availability of new data domains.
By federating and sharing knowledge, the platform provides solutions to the daily uses of data:
- Agility, productivity, and scalability of data teams,
- Control of the impacts of data on the IS,
- Support for compliance initiatives,
- ...