Table of contents for BIWA Summit 2008
- Live from BIWA Summit – Competing on Analytics
- Powering Next-Generation Predictive Applications with Oracle Data Mining (ODM)
- Critical Success Factors for successful BI and analytic implementations
- From Data Warehousing to Strategic Data Assets
- Getting to the Right Price with Oracle Data Mining
- Oracle’s BI Strategy
- Intelligent OLAP: Data Mining and OLAP
- Fraud Detection with Oracle Data Mining
Tobin Gilman of Oracle presented on Oracle’s BI Strategy. Oracle views Enterprise Performance Management and Business Intelligence as coming together over time and has a single strategy. They see ERP and CRM as having enabling operational efficiencies by driving a process-centric view across silos. Clearly BPM has completed this transition. Yet management processes – reporting, forecasting etc – are not thought of in a similarly disciplined way. Lots of different tools and approaches. Oracle sees an opportunity for management excellence layered on top of the operational excellence driven by the electronic backbone of a modern enterprise. BI/EPM is aimed at this.
Performance management is pretty well established with KPIs, benchmarks and reporting against them – even public sector organizations like NYC are doing this. Oracle sees a BI Foundation on top of which Performance Management and BI Applications are delivered within an overall EPM Workspace. To make this work you need to go against data directly although the Fusion middleware stack can also be applied. This overall stack is called Oracle’s Enterprise Performance Management System.
Oracle has recently been focused on being more “Hot-Pluggable”. For instance, any JSR 168 Portal, any database, any data mining environment, any desktop tools, any ETL product etc. Oracle clearly views this strategy as being plugged into a multi-vendor reality, not least because Oracle has built its stack by acquisition. So both Hyperion (acquired by Oracle) and nQuire (acquired by Siebel and thus Oracle) are part and parcel of the stack and both were already very open. Oracle is aiming to deliver BI, integration and analytics in the database as well as a comprehensive BI foundation.
Some functions, like data mining and integration for instance, are in both. The BI Foundation consists of Essbase (multi-dimensional OLAP), OBIEE+ (Oracle BI Enterprise Edition Plus) and Real-Time Decisions (RTD, aimed at front-office applications). Because you get questions about situations and the future (probabilities or forward-looking), need a rich platform, hence multiple components. OBIEE and Essbase have been integrated around a common enterprise information model – a semantic layer. Essbase supports planning, scenarios, profitability calculations etc. OBIEE delivers dashboards, ad-hoc query and reporting. One model is developed and used consistently reducing user support and improving compliance/upgrades. This same model is also exposed through ODBC allowing other tools to access it also. One platform, one model, multiple delivery (production reports, dashboards, desktop, proactive alerts, mobile, disconnected…). The integration allows Essbase and OBIEE to be both sources and targets for each other. Real-Time Decisions (RTD) is the third component of the platform, and for me the most interesting. It uses business rules, continual learning and predictive analytic models – Enterprise Decision Management in all but name – to make high volume, operational, customer treatment decisions. Finally 11g has a wide range of data mining capabilities (as discussed earlier).
All this platform technology is designed to support companies build what they know they need. Oracle has also developed BI Applications for many common problems like sales management or call center management. This include standard data models, plumbing, ETL, KPIs and more. Applications cover sales, services, marketing, procurement and spend, supply chain, financials and HR. These applications are designed to reduce the time to value and the total cost of ownership for these standard needs. Each consists of:
- Pre-built warehouse with defined star-schemas that integrate into a corporate model
- Pre-built ETL for large numbers of tables in Oracle applications
- Pre-defined metadata for logical models and business rules
- Best practice library of pre-built metrics, dashboards and reports
Finally there is a series of Performance Management Applications focused on the finance group. These support strategic planning, planning and budgeting, financial closing and cost/profitability management. Features include scenario modeling and simulation, impact analysis for major financial activities like debt restructuring etc as well as top-down and bottom-up planning and are based on the Hyperion product range. They share master data, business rules, assumptions etc.
Oracle has managed to bring a lot of disparate pieces into a coherent framework already. Their focus on being open and integrated is welcome and I like the number of times business rules came up in this discussion – good to see analytic folks talking up the importance of rules. I thought there was a lack of focus on decisions, however. No apparent modeling of decisions or managing of decisions against KPIs/data models for instance. I was also a little concerned that too many of these products need the same business rules and that there was nothing about how to manage these rules nor anything about how decision services, defined as part of Oracle’s BPEL/SOA stack, could use these analytics or rules. Managing decisions across the enterprise requires answers to some of these questions.