Real time business intelligence, or real-time BI, is something that a lot of companies are looking at. In addition to getting volumes of data, organizations want to getting fast access to their data and get quick insights.
Some business decisions need to be made as the information is available and seen. Consider some time critical things like fraud prevention. Waiting a week to act may mean the chance to take action will have passed.
You’ll also hear this referred to as operational business intelligence. While most business would like to have operational data available as it happens, there are several technical hurdles to overcome. Traditional business intelligence uses a data warehouse that’s optimized for the businesses reports. The reports and data are structured to facilitate quick data retrieval and easy use for the end users. The data is typically updated infrequently, most often once a day. While this is ideal for reporting and information delivery, it adds another layer of data latency and upkeep for a company’s data team.
Most transactional or operations databases are optimized for just that: transactions. They are designed to quickly be updated with new and updated transactions (sales, customers, inventory levels). They are not optimized in most cases for search or information organization, only to add or update records. Therefore, reporting against this type of database is usually slow. The additional queries also can slow down the database, and therefore the day-to-day business operations.
One solution is to have an operational data source, or ODS, that gets an off loaded copy of the transactional data many times a day. It might just get a refresh every couple of minutes. It’s not quite real-time but often times close enough. The problem here is that now you have an additional data source to manage, and it still has some lag time for data updates.
Many BI vendors and systems can report
against operational databases now. Oracle, for example, now has their
Real-Time Decisions (RTD) package. The reality of the BI landscape is
that many organizations don’t have the resources to build a dedicated
warehouse. Or, they can’t afford any large data latency in their
reporting. Understand that using real time business intelligence may
make some tradeoffs in terms of report performance and flexibility, but
it's definitely something that's worth investigating. Vitria is a newer entry into the field that is dedicated to operational intelligence.