Data warehouse architecture and design
We’ve earned our stripes delivering data implementations that are of world-class standard – from single data marts, to voluminous and complex enterprise projects – and firmly believe that a foundation of well-structured, fast-to-query, accurate data sources can make or break an organisation’s business analytics or performance management initiatives.
And when it comes to the newer concepts like data virtualisation, Hadoop, NoSQL, data lakes, massively parallel data warehouses, in-memory and the like, we can advise on the best route for your particular situation.
There are pros and cons to both the older, but proven traditional approach, versus the benefits that new technologies bring. Our iOCO team – with experience in a wide variety of RDBMS technologies, including Microsoft SQL Server, Oracle DB, Teradata, DB2, Sybase IQ and Informix – can help you find the optimal way forward.
ETL and integration
“A single version of the truth.” That’s the plaintiff cry we hear from many a frustrated business user. While this has long been the promise of BI, the real work starts much further back where data from multiple sources are integrated. Extract, Transform and Load (ETL) tools provide developers with a range of techniques to bring data together, and deal with differing data structures, business rules, granularity, timings, currencies, time-zone, languages… The list goes on and on.
We work closely with system and business analysts (either yours or ours) to determine the correct business rules, and develop robust loads to your data warehouse or other data repositories.
We have experience in Microsoft’s SSIS and IBM’s DataStage toolsets, as well as Oracle scripting skills.
It’s all about the data. That old chestnut – garbage in, garbage out – is made even more visible by the wealth of visualisation and prediction software out there. Therefore, the need to ensure that your “raw material” for all decision-making, the data, is high on many executives’ radar.
And with POPI on the horizon, securing sensitive information is in the spotlight, once again.
We tackle the following aspects of data governance:
- Data integrity (completeness, accuracy, relevancy and consistency)
Is that date a real date? Does that customer’s ID number match their date of birth? Is that product categorisation using the latest rules?
- Data stewardship and ownership (responsibility and accountability)
Do you know who to turn to when data looks suspect? Have you agreed who is ultimately responsible for the rules and quality of your data assets?
- Standardisation (single shared definitions, processes)
How can you make sure good data stays good? Do you have good processes and workflows in place to ensure that errors are picked up and dealt with in a systematic and timely manner? A visible, easily maintained glossary of business terms and rules is essential to correctly interpret data.
- Security (access to information, distribution, updates)
Protecting your data means protecting your company, your employees and your customers. Apart from the legal ramifications, it’s just the right thing to do. But finding the balance between accessibility and secrecy is an important, and ongoing, exercise.
Over the years, we have helped many companies with the data they use for decision-making. What we’ve noticed is that a business analytics, or financial performance management project, can often trigger or highlight the lack of reliable data sources.
We are well equipped to offer a managed service, either as a full outsource or to supplement your team, and can provide a number of related services, including: