To satisfy business needs regarding data integration and consumption, we

must design and develop a sound and robust Data Intelligence Platform (DIP). The reference architecture exists today.

 

The next step is to design and develop the data model for the Integration Layer in our target Data Intelligence platform.

The Data Modeller will be involved into

  • Data Modelling
  • Data Architecture
  • Metadata management
  • Data security and compliance

1      Data Modelling

This will be the core activity.

For DIP Integration Layer (DIP-IL) data modelling, we will use the usual and proven dimensional or star/schema logical modelling.

You will find thousands of pages on Internet, use keywords such as star schema, slowly changing dimension, DWH data model, dimensional modelling, surrogate/natural key, conformed dimension, Ralph Kimball

 

In our Enterprise software Sparx EA we already have a primary canvas for NLbT “Policy Transaction” Star Schema for. Attributes are not listed yet.

The BICC Data Modeller will oversee

  • Identifying the data domains and subject areas
  • Creating and maintain the DIP-IL dimensional data models into Sparx Enterprise Architect
  • Ensuring consistency and integrity across all models

We will fully decouple source system data model and DI/BI platform data model. We need to model Master Data building blocks, we need to model the conformed dimensions. These dimensions appear in our target architecture in the MDM component. We speak here about Parties, Products, Policies, Covered Items, Locations and some other concepts.

2      Data architecture

As we need to effectively load data into DIP-IL, we will need to understand where the data are available and accessible. Therefore, we need to document the data location inside IT systems or applications. This is connected to enterprise data architecture.

2.1    Data location

The BICC Data Modeller will be in charge

  • Of identifying the Golden source for any data model entity on the DIP-IL Data Model
  • Of choosing the Golden source when data are available inside several IT systems
  • Of document into Sparx Enterprise Architect

2.2    Layered Modelling alignment

The Data Modeller will ensure the alignment between the Enterprise Conceptual Data Model and DIP-IL logical data models.

The BICC Data Modeller will be in charge

  • Of keeping the DIP-IL data models aligned with the Enterprise Conceptual Data Model top-down alignment). Therefore, he/she will regularly synchronise with the Enterprise Data Architect
  • Of suggesting Conceptual Data Model upgrades following discrepancies introduced by extensions on the DIP-IL logical model (bottom-up alignment)

2.3    Metadata management

We will intensively develop data lineage inside the DIP. SAS will be the tool for metadata management: business glossary, data dictionary, data flow and transformation diagrams.

It is important to verify that metadata is aligned with Data Modelling and data architecture.

The Data Modeller will be in charge

  • Of verifying consistency between metadata and data models and data architecture
  • Of regularly publishing metadata to BICC portal accessible to the public
  • Of defining data retention rules at entity level

3      Data security and data compliance

Finally, but also important is to master access on integrated data. The Data Modeller will edict, in collaboration with data owners and future data authority, will edict accessibility and authorisations on integrated data. It is no easy exercise. Individual ownership is defined, the integration of data from several owners requires new authorizations. Also, in scope of GDPR, the Data Modeller will analyse and document procedures to fulfil rights of a subject (client or other party) and right to be forgotten.

The Data Modeller will be in charge

  • Of defining authorizations on integrated data
  • Of defining and documenting procedures to fulfil GDPR subject rights such as right to be forgotten