Developing and Deploying Data Warehouse and Business Intelligence Solutions Kerr-McGee Information Management Group Skye Brannon DW Analyst Jeff Bridgwater DW Manager Sarena Sherrard Sr DW Analyst Who is Kerr-McGee? • Kerr-McGee is an Oklahoma City-based energy and inorganic chemical company with worldwide operations and assets of approximately $10 billion • http://www.kerr-mcgee.com/ Agenda • Introduction to DW/BI Concepts • Extract, Transform & Load (ETL) • Business Intelligence / Reporting • A Day in the Life DW / BI Concepts Information Management Strategy Structure the systems and data relationships to provide user-friendly customer access to data in order to provide decision-making information Adding Value to Data Information Pyramid What is a Data Warehouse? A copy of data from one or more On-line Transaction Processing (OLTP) systems specifically structured for Query, Reporting and Analysis (QRA) • Data is typically at a summarized level to limit the size and complexity of the data warehouse • Data is usually cleansed and merged to create an “apples to apples” comparisons End-User Reporting OLTP Systems Data Warehouse The Idea Behind Data Warehousing Source Systems POS INVENTORY FINANCIALS Extract & Transform Data Warehouse Meaningful & Easy Access ? BILLING Data Information Framework Architecture Metadata Management Enterprise Reporting External Data Data Extraction Datamarts Data Cleansing Data Integration OLAP Query Data Improvement Data Mining Source Systems Data Transformation Operational Data Store Data Warehouse Project Management & Quality Assurance Operations & Systems Management Information Delivery Business Intelligence Tools Metadata Management Business Intelligence Tools • Combination of applications and External tools Data Data • Provide analysis, presentation and Extract Cleansingfacilities for users reporting and • Tailored to meet diverse needs of Integration Process executives, mgrs, analysts • Data may reside in ODS, data warehouse or data mart • Issues you choose the right tool Source – HowData Operational Data or tools? Transformation Data Store Warehouse Systems Plan/Forecast/ Analysis OLAP Query Datamarts Data Mining Information Analysis Project Management Management & & Quality Quality Assurance Assurance Project Operations & Systems Management Global / Dept/ Business Unit Summary and Analysis Data Visualization Executive Information Systems Categorize Information Needs Highly Summarized Senior Management Executive Moderately Management Summarized Business Analysts Market Researchers Integrated Operational Data Business Users Financial analysts, product managers, etc Salespersons, line managers, administrative staff, etc Information Delivery Mechanisms Wireless Mobile Web or C/S Predefined Mechanism Summaries Delivery Considerations Data Operational Trends Web or C/S Visualization Integrated with Operations? Detailed Reporting only? Real-time or based Directed on a Periodic Specialized Business CycleAnalysis (Financials) Algorithms Tethered or ‘disconnected’? C/S Standardized Web Operational Wireless Reporting Operational Trends (in millions) Net Revenues Net income Earnings per share Ad-hoc Queries Ret urn on net revenues Cash & s/t investments Tot al Assets Shareholder Equit y 1998 1999 $x,xxx $x,xxx xxx xxx 2000 $x,xxx xxx 2001 2002 $x,xxx $x,xxx xxx x,xxx x.xx x.xx x.xx x.xx xx% xx% xx% xx% xx% $xxx $xxx $xxx $xxx $xxx x.xx $xxx $xxx $xxx $xxx $xxx xxx xxx xxx xxx x,xxx Web or C/S Operational Trends B.I Infrastructure All things Cognos - Cognos is a vendor The suite of applications we’ve bought and use from Cognos are: - Access Manager – (Security) - Upfront – (Portal – http:/intranet/kmbi) - PowerPlay – (reports/cubes) - Impromptu – (Web based PDF reports) - NoticeCast – (conditional report notifications) - *Visualizer – (graphic depictions of data warehouse/cube information) *What we will cover Terminology Cube - A multidimensional way to analyze information, designed to provide quick answers to the who, what, why, when, and where business questions Drill Down - Going from a summarized view to a more detailed view of information within the same cube Drill Across - Linking data from One Subject Area to Another (General Ledger to Accounts Payable) Drill Through - Linking to source data using selected filters Powerplay Web - On-Line Analysis Tool for cubes (slice/dice, drill down, drill across & drill through) Newsbox -A web based folder used to store views of data (reports) Every KMBI user has their own personal newsbox Cognos - Upfront - Upfront - Portal Management Cognos - PowerPlay - PowerPlay – web reports/slicing and dicing/data analysis, based on cubes More Information on Cognos website: http://www.cognos.com/products/businessintelligence/analysis/ Cognos - Impromptu - Impromptu – printable reports (in PDF) that may/or may not be produced with prompts for filtered information Cognos - Visualizer - Visualizer – interactive graphic depictions of data warehouse/cube information Developing Visualizations • Initial Project meeting should include: – Client - gives input on look and feel, data requirements, timelines – Project Manager – ensures project is feasible within budget and time restraints at the onset and through out the project – Data Warehouse Architect – ensures all the needed data is in the data warehouse – Business Intelligence Specialist (cube builder) – ensures all the needed data is in the cube, in the correct format – Web Interface Specialist (visualization builder) – works with cube builde and client to established look and feel, navigation, chart styles, etc Day in the Life Typical Day