Success story Randstad

Identified needs

Randstad Spain's data office needs to acquire a data governance platform that will enable it to operationalise the governance model for the company's data assets, accompanying the data management programme as an asset at all times (Data Driven Strategy).

  • The selected platform is Anjana Data. The main requirements it must meet include:
  • Adaptability to the already defined governance model and metamodel.
  • Time-to-market of 4 months as a strategic objective.
  • Integration with the Omma Data Quality tool.
  • Be autonomous in the functional and technical management of the platform. Make it scalable in all key aspects.

Use case

The governance model proposed by Randstad Spain has been in development for some time, using office automation tools and from a theoretical perspective. It is now at a turning point, requiring a platform to implement everything that has been defined so far.

This definition centres on the concept of business terms. Far from being a simple functional and agreed definition of business terminology, this concept underpins and is central to Anjana Data:

  • Classification and relationship with technical assets, including relationships with:
    • Datasets hosted on operational systems.
    • Datasets present in the various phases of analytical systems and data warehouses, centralised in a Data Lake, whose core technology is GCP Bigquery.
    • Technical solutions, including data services offered to business users in the form of dashboards, reports, etc., which are the result of analytical and BI operations.
  •  The applied business ontology, both in Spain and internationally, in terms of which terms are equivalent in nationally defined terms and their international definitions.
  •  Data quality, at the business term level, is the axis around which information quality is controlled. That is why Anjana Data has implemented quality-related information in both technical assets and the business term itself. Furthermore, in its integration model with the Data Quality tool, Omma Data, an integration model has been designed in collaboration with the integrator Mática Partners that centralises and displays quality information in the business term.
  •  The very definition of the governance model, which is currently based on the definition of the organisation, allowing for the agile definition of those responsible for it at all levels, including key aspects such as quality, safety, technical responsibility and functional responsibility.

To achieve all this in a first use case, intensive work has been done on the definition of the Anjana Data configuration itself in order to have a fully customised configuration for cataloguing the first objects of the metamodel in Anjana Data.

The selected use case is based on a small part of the data lake and all the objects and relationships in the metamodel that are related to it. To this end, certain data services have been selected, which are actual dashboards used by the business in its day-to-day operations, and have therefore been catalogued, including:

  • Data solutions: Dashboards.
  • Datalake datasets originating from dashboards (using Anjana Data's native connectors with GCP Bigquery technology)
  • Business terms that encompass Datasets.

At the same time, an initial governance model has been implemented, based on roles and permissions, which will evolve over time alongside the data management programme.

 

Objectives set

In response to the needs identified, the objectives have been defined directly in line with these requirements:

  •  Be fully configurable and adaptable: so that the platform adapts to the governance model and the current state of the data-driven strategy, and not the other way around. At Randstad International and Randstad Spain, we have worked to develop a governance model that revolves around business terms, which is why the platform must be fully configurable in order to transfer everything that has been defined and worked on so far as directly as possible, both in Spain and internationally.
  •  Minimum time-to-market. Randstad Spain strives for effectiveness, so the implementation of the tool and development of the first use case, in order to see results, must be as short as possible.
  •  Integrated with the Omma Data Quality solution. The Anjana Data implementation project is running in parallel with the Omma Data implementation, which is why we want to implement an integration mechanism between the two tools.
  •  Be autonomous in the functional and technical management of the platform. During the implementation project, Randstad Spain wishes to achieve sufficient autonomy to be able to administer the Anjana Data platform from both a technical and functional point of view.
  •  Scalable over time and across the volume of data assets to be governed. One of the objectives of the Data Driven strategy is to be able to govern the entire Data Lake, so the platform implemented must at all times support both the functionalities and the architecture and licensing of both the strategic data management programme phase and the company's data management maturity status.

 

Functional and technical architecture

 

Benefits obtained

Taking into account the needs, objectives, and use case selected for development at Anjana Data, an implementation project has been designed to cover all of these aspects:

  •  Adaptability: Randstad Spain has its own configuration of Anjana Data that is fully adapted to the previously defined governance model and metamodel. Customised entities have been configured, such as Business Term and Data Product, among others, and at the same time, native entities offered by Anjana Data have been configured, such as Dataset, DatasetField, processes, instances, and solutions, so that they contain the metadata (characteristic information) that allows these elements of the Randstad model to be governed effectively in all desired aspects, such as custody, quality, security, lineage, etc.
  •  Time-to-market: During the four-month implementation project, the goal of having an Anjana Data instance with tangible information showing the data governance information for the chosen use case was achieved, serving as a lever to drive and defend Randstad Spain's determination to become a data-driven company following best practices in data management.
  •  DQ Integrations: Through an incremental model of ad hoc integration between Anjana Data and Omma Data, information corresponding to data quality can be viewed in Anjana Data, not only at the dataset level, as proposed by native integration, but also at the business term level. This adaptability has made it possible to maintain the level of custody intact as these two tools have been introduced into the Randstad Spain data ecosystem.
  • Autonomy: Thanks to Anjana Data's three main accelerators: official certifications, guided workshops included in the implementation plan, close collaboration between Anjana Data's Customer Success team and Mática Partners' integration team, and the stakeholders involved from both the data office and Randstad Spain's IT team, the latter have sufficient knowledge to independently handle administration, maintenance, and decision-making regarding the next steps with Anjana Data.
  • Scalability: Anjana Data's modular architecture facilitates both horizontal scaling of its components, allowing it to support a higher volume of concurrent usage than originally estimated, when necessary. At the same time, Anjana Data's licences, based solely on the number of governed objects, will allow Randstad Spain to increase the number of users, the number of professionals certified in the tool, the use of new and future connectors, and access to new versions under the same licence.

Challenges overcome

Although the initial implementation of a governance model and metamodel in a company is a major challenge in itself, certain individual challenges arose during the implementation project that were successfully overcome:

  • Adaptability of the platform to what has been previously defined and worked on. Anjana Data can be defined in a way as a blank canvas on which to draw the entire governance model and metamodel that best suits the organisation's needs with regard to information management. To this end, Anjana Data's integrator, Mática Partners, has worked conscientiously with both Randstad's Data Office to understand and analyse what had been defined and with Anjana Data's Customer Success team to analyse, design and propose solutions so that the configuration of Anjana Data fully meets both the needs and expectations of Randstad Spain.
  • Display quality information at the business level. The proposal for native integration between Anjana Data and Omma Data is based on Anjana Data's native Dataset object and Omma's Quality Point. this did not meet Randstad Spain's expectations and needs, as the governance model focuses on the Business Term, which, in terms of information quality custody, is associated with the quality values of the datasets it represents. Once analysed, Mática Partners, thanks to the API integration capabilities of the two tools, implemented a first phase and an iterative plan to achieve complete and automated integration that would allow quality information to be obtained, offered and governed at the business term level, as designed in Randstad Spain's original governance model.
  • On-time implementation project. The successful implementation of Anjana Data's plan, meeting all deadlines and objectives, is undoubtedly a shared success between Randstad Spain's data office and IT department, the Mática Partners integration team, and Anjana Data's Customer Success team. Direct collaboration, persistence and know-how have enabled smooth, consistent work and agile decision-making, resulting in a use case being implemented at Anjana Data, among the other objectives mentioned above.
  • Governing data in a multi-cloud environment. In the case of Randstad Spain, Anjana Data has been deployed on one cloud provider, while the data to be governed from the data lake is on another provider. This has not posed any problems, as Anjana Data connectors only need a valid credential and an open communication channel to extract metadata from them. The Randstad Spain Infrastructure team has made the necessary arrangements, and Anjana Data is now connected to data sources outside the cloud where it is deployed.