Archive for category SAP HANA and Analytics
Attended a seminar focused on SAP HANA Predictive Analytics RDS @ SAP Labs today. Came back impressed with what I saw. This RDS comes loaded with a lot of pre-built predictive analysis models for various industry verticals and for various business scenarios. While PAL on HANA is an intermediary tool still requiring technical knowledge, this offering has a user-friendly front-end which is easier to navigate and use. Version 2 of this product is due for release sometime next week. What are some of the cool features I saw and liked:
1. Has pre-built models for 5-6 industry verticals
2. Easy to use UI for deploying the predictive analysis capabilities
3. Built-in visualizer using HANA visualizer and BobJ Explorer
4. Accesses the HANA Live calculated views for information (single source for actuals and forecast)
The next step is to further enhance this product using the new SAP acquisition KXEN (renamed SAP InfiniteInsider)
Good to see SAP making such huge progress in the predictive analysis space; I had to trya and make BobJ Pioneer (don’t even know what it is called now), go head-to-head with the likes of QlikView; even Lumira was not quite there. What I saw today gives me great hope that customers can stay within the SAP ecosystem and get most of their business IT needs met.
Of late, as I have been doings the rounds with our clients and prospects talking about data analytics strategy with SAP HANA, and talk about the Enterprise Data Warehouse, I am left to wonder if the EDW as we all know it is dead (or at least dying). SAP HANA provides businesses the capability to build report, charts, and dashboards using pre-delivered analytic and calculated views with minimal help from IT. Anybody with some basic knowledge of front-end tools like SAP Business Objects can directly access these views and create their own reports (help.SAP.com has links to the pre-delivered content).
The pre-delivered views along with the capability of HANA to aggregate data from non-SAP sources like Hadoop, makes this a powerful tool for the business to do their own aggregation based on their requirement.
Of course one could argue that SAP BW with its standard extractors had similar promise; but I think SAP BW was a different beast. Self-service in SAP BW was just not that simple. SAP HANA’s pre-delivered analytic and calculated views are a lot easier to access and use by a business user savvy in the use of analytical tools.
So where does this leave the Enterprise Data Warehouse, which in the SAP world used to be SAP BW; SAP continues to maintain that SAP BW is not being replaced by SAP HANA and SAP BW should be maintained as the semantic layer for data governance and management, essentially the EDW solution. While this may hold true for companies that have invested significantly in SAP BW, it would be difficult for an organization that does not currently have a EDW strategy or does not have a significant BW presence to try and build an EDW strategy from scratch using SAP BW. It is not the lack of capability- it is just that an EDW strategy takes time and building this in SAP BW does not make this any easier.
I was trying to explain this to an IT director at a client and he indicated that he expects the business to give the requirements to IT and it is ITs responsibility to provide them the information. This I think is simplistic thinking at best; unless you have an organization which has a very strong push from the top to only use information coming from IT, this strategy will be difficult to maintain and will lead to multiple data marts and solutions popping up all over the place.
The best strategy for IT is to maintain control over regulatory and compliance type reports which demand and require accuracy and ‘one version of truth’ like SEC filings, EEOC reporting, etc. It is easy to get the required support to channel this kind of information from a central source owned by the business, but maintained by IT and with strong governance and controls.
Operational reports are best left to giving the business access to the data and not spend time trying to put a lot of control and governance in how they are used.
This allows IT to be viewed as an enabler instead of being a roadblock to information access.