An Update – SAP BI and EIM 4.0

It’s been a while since a major release of BusinessObjects/SAP BI so 4.0 is a big deal – over a million person hours invested. SAP’s overall focus is to help companies “run smarter” and obviously the work on BI 4.0 has to be part of this overall environment – how do analytics help companies run more effectively, more consistently, more efficiently.

It’s been a while since a major release of BusinessObjects/SAP BI so 4.0 is a big deal – over a million person hours invested. SAP’s overall focus is to help companies “run smarter” and obviously the work on BI 4.0 has to be part of this overall environment – how do analytics help companies run more effectively, more consistently, more efficiently. With 4.0 SAP took a broader look than just the BusinessObjects portfolio – this release pulls together the whole portfolio of Business Analytics products at SAP along with reference architectures and analytic applications (largely pre-defined content, dashboards and reports for application areas).

Running smarter to SAP means Knowing your business better, Deciding with confidence and Acting boldly. The big themes in this release are:

1. Big Data in Real Time
2. Insight with confidence across business and social data
3. Instant, powerful BI in your hands
   f course this seems to mean BI on high-end mobile devices rather than pushing BI into the hands of people who don’t have those devices or any interest in using BI tools. Ho hum.
4. Best fit, right size, right now

The key element in the big data in real-time is SAP HANA – High Performance Analytic Appliance. This pre-configured hardware/software stack is designed to support in memory analytic processing based around a columnar database. Obviously the size of blades available has increased steadily while the cost of memory has dropped dramatically, changing the dynamics for in-memory processing and pushing this towards being a pervasive approach more quickly than you might have thought.

HANA calculates everything on the fly, avoiding the need for specialized/localized tuning, and is designed to support the SAP Business Suite, SAP Netweaver BW, SAP BusinessObjects and other interfaces for third parties like SQL. It is designed to provide real-time analysis and be a flexible foundation that minimizes data duplication. And while this supports integration with SAP Netweaver BW (the data warehouse offering from SAP), companies could commit to HANA only and use BusinessObjects directly against it. Companies committed to SAP Netweaver BW can use the BW Accelerator and those doing non-SAP data could use Sybase IQ, all with the BusinessObjects stack on top. They expect over time for the various products to collapse down to a more streamlined suite that still supports these three use cases. Interestingly there is some work around FuzzyLogix too along with some thinking about when it might make sense to bring this kind of in-database analytics to the in-memory offering.

New SAP Business Objects Universe design tool gives the Universe two separate layers working together – a data foundation layer and a business layer. The separation of perspectives allows data architects to work with the potentially very complex data sources while the business analyst can work with this to define a business vocabulary. The data foundation layer has been worked on to make it more scalable and to handle multi-source data more easily – you can build a single Universe across a federated data set. It also handles multi-dimensional data more effectively. Once the Universe is defined it can be easily used within things like the dashboard designer.

Now today this does not allow either predictive analytic outputs to be rapidly integrated into these Universes nor can models be built against these Universes. Similarly the rules engines (BRFplus and Netweaver BRM) cannot access the Universes directly. Both of these are recognized as important scenarios, however, and SAP is working on integrating Universes in these ways. This of course is excellent news as a coherent semantic layer across reports, dashboards, predictive analytics and business rules is a great platform for Decision Management.

The new Information Steward product has some interesting features but it sadly introduces a new rules format for data quality rules. As many of these rules will also be used in checking transactions for completeness and validity in the applications it would have been nice if these could have been reused. Over time there is a clear desire on SAP’s part to reduce the need for rules to be written in multiple formats – what they need (and what I hope they will build) is some good federated rule management so that you can apply common rules in multiple contexts. After all, as I always tell people, rule management will trump rule execution in the end.

Lots of interesting stuff.