Download to read offline
Uploaded byAli BELCAID
Albel Pres Continuous Intelligence Overview
The document discusses how traditional business intelligence (BI) architectures can be improved with a business event approach. It notes that traditional BI relies on past data and is not synchronized with business processes. The new approach involves capturing real-time business events, processing them using streaming analytics, and taking immediate action. This allows businesses to gain insights from current operations and respond quickly to issues or opportunities without data latency. Implementing event-driven BI brings real-time support for business decisions.
Albel Pres Continuous Intelligence Overview
- 1. Event Driven BusinessIntelligence By Ali BELCAID – Managing Consultant
- 2. Context : TraditionalBI New BI Architecture Business Event Approach to Project Business Event Implementation Event Management - Where it is Applied
- 3. Context Traditional BI Architectures Business Intelligence (BI) ERP • Reports • Olap • Dashboards CRM ETL ETL FIN ETL Datawarehouse or ODS HR Files Real-time BI Operational sources Datamarts • Real-time Operational • Relational Databases Dashboards • Mainframe • Alerts and escalation • Flat files • Collaboration
- 4. Context Traditional BI Architectures As matter of fact : Business users are relying on inflexible, heavy and costly BI platform : each time they ask for new request, they have to wait for Business users are relying on past figures, so that they are delaying their decision every time waiting for the right information until they get it Their BI platform is not synchronized with their business processes : there is a clear GAP between their BI and their processes There is no way to have insights from what Business User is doing at the moment : He has to wait for having it ! Many companies are struggling to get people having real-time data but they are doing it in the wrong way
- 5. Context Continuous Intelligence – Avoid Latency Data Latency is what is causing Traditional BI not to reach the Business Event goal of having the right data at Trigger every Data the right time. Handling Do not wait for a long Time: Business Event quicker response to detected issues Potential Business Value Data Stored Avoid : Information Delivered Data Latency Analysis Latency Decision Latency Information Analyzed Action Taken Action time
- 6. Context Real World Event In real world (and real business also) Customer Logon we are surrounded New by Events. We have Order to respond to these Customer events in such way Checks Fed “Close Account” there will be no Base Rate Web Page interruption in what Increase we are doing : This Contract is where Business Production Mobile Call Submitted Event comes to play Item from CT New @11.13 a new role in the BI Arrives at Liability universe as support Store Added Contract to our Business Returned decisions. Rental thru EDI Rental Car Car Returned Crashed
- 7. Context Where is the information ? These Events when Risk of there identified they Fraud Customer Change in Risk! Defection Product Sales could mitigate many risks that companies Trend are facing in day to Rental day operations Contract Customer Cell phone Complete CrossSell fraud alert Opportunity Employee Over hours Contract Compliance Validated Limit Approached Customer now rated Gold Contract Valid So, applying Business Event Processing to BI makes a real sense to bring real-time support decision to Business Users
- 8. Context : TraditionalBI New BI Architecture Business Event Approach to Project Business Event Implementation Event Management - Where it is Applied
- 9. New BI Architecture Streaming Analytics Store Data for : Trend Analysis Business Intelligence (BI) Comparative analysis Reports Analysis on historical data Olap ERP ETL Dashboards ETL CRM ETL Datawarehouse or ODS FIN ESB A HR D A Filter / Aggregate Datamarts P Files T Capture Business Event Correlate E Processing X R Event Real-time BI Operational Dashboards S Events Act Database Alerts and escalation Collaboration External Data Continuous data flows
- 10. New BI Architecture Example – Action from Insight This is an illustration of the way a Marketing Director can manage his Execute Action campaigns according Provide values for the following action parameters: to info's brought into Begin Marketing Campaign his dashboard. This is Region West Campaign ID 42684 all about bringing intelligence into processes. Help Cancel OK Bringing real-time data will turn this into Continuous Process Intelligence
- 11. Context : TraditionalBI New BI Architecture Business Event Approach to Project Business Event Implementation Event Management - Where it is Applied
- 12. Business Event Approach How do we tackle a Business Management Event Project? Bringing Event Processing Management into BI world requires the use of new approach to tackle projects. You can use a phased approach containing 3 steps : Sense, Analyze and Respond. This approach is well known in the CEP (Complex Event Processing) arena. Sense Analyze Respond Event 1 Event 2 Correlation with external Events - Alert in Dashboard Event 3 Act by - Activate a Process Business Rules Making - Activate another application Event 4 Decision -… Event 5 Analysis … The analysis part correlates or The responders execute the fuses data to determine an then clauses. This part Event n estimate of the current state, and responds when reality to compare the real state with deviates from expectation by the plan. In effect, the analysis modifying expectation (the The sensing part part continuously evaluates the plan) and executing obtains information when clauses of when-then responses such as sending from within and outside rules by fusing information from alerts, invoking applications, the enterprise. This all the sensors and determines if and initiating operations by part obtains the data the then clauses should be actuators. that defines “reality.” executed.
- 13. Business Event Approach Project Phases - Sense Functional Requirements Events 1. Determine Business Objectives that you are looking at 2. Determine the kind of events you want to manage and if your system allows you to get them internally or externally. Group 1 3. Classify and group events that could be threats or opportunities depending on the expected analysis. Group 2 Technical Requirements Group 3 1. Determine Event Sources • Do you have access to the events ? Group 4 • What are the security Requirements 2. Determine Event Transport Services • Publish and subscribe messaging • Feed backed transport such as RSS or ATOM • Request / Reply (Pooling) and SOA 3. Determine your adaptation requirement • SNMP • File adapters • Database
- 14. Business Event Approach Project Phases - Analyze true true Group 1 false false 1. What are the 2. What is the sources of temporal or spatial 3. What rules & calculations should be applied? data? link between the data sequences? Complex Event match Stream1 match Stream 2 Stream 3 4. How to correlate ? Stream n Stream Adapters Event Correlation
- 15. Business Event Approach Project Phases - Respond Decision-Centric Automation Investigate Manage to Respond Routes work among Exceptions people Keeps track of work status Task-oriented Collaborative Informational Automated Transaction- Detection Centric Automation Straight-through- Automatic processing Response Optimized to avoid manual steps and redundancy Event-driven
- 16. Context : TraditionalBI New BI Architecture Business Event Approach to Project Business Event Implementation Event Management - Where it is Applied
- 17. Business Event Implementation TIBCO Business Event Server – The platform Complex Event Processing Event Filtering, Aggregation, Correlation Missing Event Detection Causal & Timing Relationships Multiple Event Patterns & their relationships Reactive Rules Triggered By Event Patterns End-to-end visibility into corporate performance SLA violation Fraud detection Up-sell opportunity Offers Real-Time Analysis Source Tibco® of Events
- 18. Business Event Implementation TIBCO Business Event Server – State Model – Visualize Object Life Cycle Source Tibco®
- 19. Business Event Implementation TIBCO Business Event Server – BE Features – Rules Analyzer • Visualize dependencies of Rules. • Eliminate conflicts among Rules/Conditions. • Visualize Concept and Event relationships through rules. • Drill down to a concept, and see its dependencies. • Visualize what gets fired when something changes, and its ripple effect. • Profile runtime metrics/statistics of the engine for each rule, and optimize the rule. • Combined with the debugger, shows when a rule is fired. Source Tibco®
- 20. Business Event Implementation TIBCO Business Event Server – BE Features – Decision Manager Source Tibco®
- 21. Business Event Implementation TIBCO Business Event Server – BE Features – Rules Tester Source Tibco®
- 22. Context : TraditionalBI New BI Architecture Business Event Approach to Project Business Event Implementation Event Management - Where it is Applied
- 23. Business Event Management Fields where it is applied Financial Services Health Care Telecommunications Algorithmic Trading in many asset Patient monitoring Network management classes Fraud detection SLA management MIFID compliance, OSS/BSS systems Surveillance Location-based services Public Sector / Government Fraud detection SWIFT money transfers Electronic battlefield Retail / Supply Chain Real-time risk management Surveillance Smart order routing Emergency response Automated supply chain Market making War fighter monitoring Smart shelves Market aggregation Security Real-time point-of-sales analysis Treasury centralization Net centric applications and one-to-one marketing International funds transfers ATM monitoring and control Web Clearing Travel Cash and securities payments Click-stream analysis with interbanking compensation Flight logistics and operations Real-time web page analysis Automated customer service On-line fraud detection Energy (kiosks, etc.) Automated and computer Entertainment Energy trading assisted baggage handling Pipeline monitoring (e.g., with RFID) Casino fraud detection Power grid monitoring and Real-time fare pricing Dynamic customer loyalty programs control On-line gambling and gaming.
- 24.