Commercial Data Warehouse on Teradata
This project started in mid-year of 2015 and is ongoing. In Phase 1, it aimed at transitioning of the existing legacy data warehouse powered by Oracle over to Teradata to realize cost savings. The goal was not simply “lift and shift’, but to replace the legacy solution with the one that meets the current business requirements, enriches the functionality and is able to adapt to the future needs.
Commercial data warehouse has been implemented on Teradata using DataStage and Mulesoft as ETL and QlikSense as a front end. Architectural tasks included the development of the foundational standards and a clear delivery model to ensure unified approach to ETL build and front-end consumption. The implemented data load strategies, data processing routines and control mechanisms ensured smooth integration of the existing data and paved the path for future additions. We also developed data validation processes, including exception views and data refresh dashboard, supporting daily monitoring. All of it helped to restore the business trust in data. The standards and loading processes are thoroughly documented and serve as source of valuable information for operational support.
An existing data management system is an integrated component of the overall solution. It is built on Kalido MDM. In fact, master data are sourced from two MDM implementations (both Kalido based), given the transitional state of the parallel MDM consolidation project. Similarly, the standards and boundaries have been developed to ensure smooth master data flow and, at the same time, efficient transition to the future consolidated MDM.
As this warehouse program proceeds, ensuring governance and conformance to standards will be a key consideration.
Our involvement continues in the area of the extensions to the existing warehouse, including new sources, subject areas and capabilities.
Enterprise Data Warehouse Solution on Teradata
In this ongoing program, we act as the data warehouse lead on the core team of a new enterprise data warehouse solution on Teradata using Kalido DIW, Informatica and Cognos. This involves the overall architecture, modeling, and DW build responsibilities, as well as ensuring that the meta data, governance and security processes are clearly established. The first two subject areas in the cross functional warehouse were delivered at the start of 2012, with five further subject areas added later in the same year.
As this warehouse program proceeds, ensuring governance and conformance to standards with multiple concurrent project teams will be a key consideration, as will ensuring data security and restricting access to sensitive data, not only to end consumers, but also across the multiple support and development teams.
Financial Data Warehouse
This project, a financial data warehouse implemented with Kalido on Oracle using a Cognos front end, was delivered with our help in 2007.
- DM Review Innovative Solution Award
- Ventana Research 2008 Information Technology Leadership Award
- 2007 InfoWorld 100 Awards (Sales and Financial Data Warehouse Initiative)
Our involvement continues in two areas:
- Extensions to the existing warehouse, including new subject areas
- Analytical Master Data Management solution (using Kalido MDM), to support hierarchy management and integration of data from disparate data sources.
Global Federated BI Solution
With this CPG customer, we act as the overall architect and take on a governance role in the design and implementation of a global BI solution, involving both a data warehouse build and supporting master data solution. There is a core foundation of the model and related build that is ‘global’ and shared across all regions. The actual implementations are regional, allowing extensions to the core model to support business differences. The foundation is based on Kalido DIW and MDM on Oracle, with Informatica used for ETL. Currently both Cognos and Business Objects are used for BI reporting, and it is possible that some regions may use an alternate ETL tool, and potentially Teradata rather than Oracle depending on data volumes.
The BI solution involves multiple subject areas with shared core hierarchies, allowing the possibility of cross functional analysis. It is expected to expand both geographically and functionally in the next year. A complicating factor in the implementation is that different partners are used to support development in different regions and across different subject areas. Governance and security, as well as assuring adherence to standards, is a key consideration.
The foundation of the model for this solution is based on the completed Asia / Pacific BI referred to below. The actual initial build is for the European region but will be back ported to the Asia implementation and then extended as required for use in other regions.