The following slides were developed while consulting at 12 of top 25 world banks about financial systems from 2013-2016. Although not specific to GenevaERS, it delineates the causes of the problems in our financial systems, and the potential impact of GenevaERS supported systems and the Event Based Concepts.
The recent financial crisis has exposed the systemic problem that the world’s largest financial institutions cannot adequately account for and report on liquidity, positions, currency exposure, credit, market, and interest rate risk and product, customer and organizational performance. The CFO plays a critical role in correcting this problem by leveraging the financial data they already control, as well as leveraging scale to take out cost. But even industry insiders do not realize that financial institutions suffer a unique set of domain problems when it comes to financial reporting.
Current financial reporting systems are antiquated and very inefficient. They were designed decades ago to simply track flow of capital, revenue and expenses at the company and departments levels. The lack of transparency is evident in the increasing costs of the finance function with few benefits to show for the investment. Sarbanes Oxley and other regulations have proven ineffective at getting at the root of the problem and the resulting financial meltdown regulations may well prove similarly ineffective. These pressures create diseconomies of scale which affect the largest institutions the most.
For the most part, existing systems deliver accurate results in summary, but the increase in transparency requires line of site to the underlying causes of those results. Consider if your personal bank account statement or credit card bill only presented the change in balance between periods, but provided no list of transactions. When the statement is as expected, further detail may not be needed. But when the balance is in question, your first response is ‘why’ and you immediately want to see the transaction detail. The same issues are at stake when managing the finances of the enterprise – with the associated cost and consequences considerably higher! A single instance of financial restatement has cost some organizations hundreds of millions of dollars to correct, not counting lost market valuation.
Currently 90% of the money supply in mature markets is represented by digital records of transactions and not hard currency. It’s no wonder that that the volume of electronic finance records being kept has exploded compared to when the systems were first created. Yet our approach to these demands has not been to automate the process of keeping and accessing the details of the transactions. Almost all employees in today’s financial institutions are involved in capturing and coding financial details in some way, and a large number of non-finance employees are involved in the investigative process to locate the additional detail so often required. The effort for this manual intervention is incredibly inefficient and costly.
As we see all around us, computing capacities have increased by several orders of magnitude since these finance systems were designed. However, reporting processes have grown organically as a system of transaction summaries in order to continue to bridge multiple financial systems – but have lacked a single unified approach. This has meant that for the most part the business of financial reporting has not benefited from the increase of computing capacities available today.
A Smarter Planet is founded on financial markets that provide for greater transparency and comprehension of the financial reporting by bank and non-bank entities, allowing the markets to react to conditions in more informed, less reactionary ways. IBM has spent 25 years refining an approach to this for financial institutions. The IBM® Scalable Architecture for Financial Reporting™ (SAFR) system provides financial reporting that is built bottoms up from the individual business event transactions to provide the finest grained views imaginable.
By harnessing today’s computing power and straight through processing approach, the details behind summary data can be made available in seconds rather than days or weeks. Providing nearly instant access to the highest quality financial data at any level of granularity will eliminate the duplicative reporting systems which tend to capture and produce summaries of the same business events for many stakeholders and reporting requirements.
More importantly, it will automate the hidden work of armies of people who are required to investigate details and attempt to explain results, or attempt to reconcile the disparate result of these many reporting systems—a truly wasteful activity caused by the systems themselves. Keeping the details in a finance system that can serve these needs allows for increased control, quality and integrity of audit efforts rather than dissipating them.
Some may question how much detail is the right level of detail? Others may suggest this is too radical a change in a mature, understood and tested set of systems. IBM experience with some of the largest financial services companies suggests that building a finance system. based on the requirement to instrument the most granular level of transaction detail immediately stems the tide of increasing costs, lowers a variety of risks and can be a key driver of transformation of the banks ability to become more agile. In time this approach begins to provide economies of scale for reporting.
SAFR is: (1) an information and reporting systems theory, (2) refined by 25 years of practical experience in creating solutions for a select group of the world’s largest businesses, (3) distilled into a distinctive method to unlock the information captured in business events, (4) through the use of powerful, scalable software for the largest organization’s needs, (5) in a configurable solution addressing today’s transparency demands.
Companies expend huge sums of money to capture business events in information systems. Business events are the stuff of all reporting processes. Yet executives report feeling like they are floating in rafts, crying “Data, data everywhere and no useful information.” Focusing reporting systems on exposing business events combinations can turn data into information.
Although analysis of business events holds the answers to business questions, they aren’t to be trifled with, particularly for the largest organizations. Reporting processes—particularly financial reporting processes—accumulate millions and billions of business events. In fact, the balance sheet is an accumulation of all the financial business events from the beginning of the company! Such volumes mean unlocking the information embedded in business events requires fundamentally different approaches. The 25 years of experience of building SAFR in services engagements has exposed, principle by principle, piece by piece, and layer by layer the only viable way.
This experience has been captured in a method of finding and exposing business events, within the context of the existing critical reporting processes. It uses today’s recognized financial data like a compass pointing north to constrain, inform, and guide identification of additional reporting details. It facilitates definition of the most important questions to be answered, and to configuring repositories to provide those answers consistently. It also explains how to gradually turn on the system without endangering existing critical reporting processes.
The infrastructure software, a hard asset with hundreds of thousands of lines of source code and feature set rivaling some of the best known commercial software packages, is most often what is thought of when someone refers to SAFR.
The Scan Engine is the heart of SAFR, performing in minutes what other tools require hours and days to do. The Scan Engine is a parallel processing engine, generating IBM z/OS machine code. In one pass through a business event repository it creates many business event “views,” providing rich understanding. It categorizes, through join processes, the business events orders of magnitude more efficiently than other tools. Built for business event analysis, it consistently achieves a throughput of a million records a minute. It is highly extensible to complex problems.
SAFR Views are defined in the SAFR Developer Workbench or rule based processes in the SAFR Analyst Workbench or in custom developed applications. The Scan Engine executed as a scheduled process, scans the SAFR View and Metadata Repository selecting views to be resolved at that time.
The Indexed Engine, a new SAFR component, provides one at a time View resolution through on-line access to Scan Engine and other outputs. It uses Scan Engine performance techniques. Reports structure and layout are dynamically defined in the Analyst Workbench. The Indexed Engine creates reports in a fraction of the time required for other tools. Its unique capabilities allow for a movement based data store, dramatically reducing data volumes required both in processing and to fulfill report request.
Upon entering Managed Insights, users select parameters to drill down to increasing levels of business events, and perform multidimensional analysis through the Viewpoint Interfaces. The Insight Viewer enables discovery of business event meaning in an iterative development mode.
The SAFR Infrastructure Software has been configured over 10 years for number of clients to provide an incredibly scalable Financial Management Solution (FMS) for the largest financial services organizations.
The heart of FMS is the Arrangement Ledger (AL). An “arrangement” is a specific customer/contract relationship. The AL, a customer/contract sub-ledger, maintains millions of individual customer/contract level balance sheet and income statements. This incredibly rich operational reporting system supports a nearly unbelievable swath of information provided by scores of legacy reporting systems in summary, with the added benefit of being able to drill down to business event details if needed. Doing so allows reporting high quality financial numbers by customer, product, risk, counterparty and other characteristics, all reconciled, audited, and controlled.
AL is fed daily business events typically beginning with legacy general ledger entries and then transitioning to detailed product systems feeds over time. The business events become complete journal entries at the customer-contract level, including reflecting the impact of time in the Accounting Rules Engine. Rules are under control of finance rather than embedded in programs in source systems, enabling Finance to react to changes in financial reporting standards, including International Financial Reporting Standards (IFRS).
The business event journal entries are posted by the Arrangement Ledger on a daily basis, while it simultaneously generates additional point in time journal entries based upon existing balances, including those for multi-currency intercompany eliminations, GAAP reclassification and year-end close processing. It accepts and properly posts back-dated entries to avoid stranded balances, and summarizes daily activity to pass to the General Ledger. The General Ledger provides another control point for the traditional accounting view of the data. The Arrangement Ledger detects and performs reclassification keeping the arrangement detail aligned with the summary General Ledger
AL also accepts arrangement descriptive information with hundreds of additional attributes to describe each customer-contract, and counterparty or collateral descriptive attributes, enabling producing trial balances by a nearly unlimited set of attributes, not just the traditional accounting code block. Extract processes produces various summaries, perhaps ultimately numbering in the hundreds or thousands, to support information delivery for not only traditional accounting but also statutory, regulatory, management, and risk reporting. The SAFR one pass multiple view capability allows AL to load data, generate new business events, and create extracts all in one process, including loading the incredibly information rich Financial Data Store.
Information Delivery includes multiple ways of accessing the Arrangement Ledger and Financial Data Store. The major window is through SAFR Managed Insights. This parameter-driven Java application provides thousands of different permutations of the data. It allows drill-down from summaries to lower and lower levels of data without impacting on-line response time. It allows dynamic creation of new reports and multi-dimensional analysis of Financial Data Store data. Extract facilities provide the ability to feed other applications with rules maintained by finance. Other reports provide automated reconciliation and audit trails.
FMS can be tailored to work within an existing environment, including working within the existing security and reference data frameworks. FMS is often can be a sub-component of an ERP implementation.
This is a financial system architecture for the 21st century. This is the reporting system architecture for the 21st century. Finance transformation starts with finance systems transformation. Finance systems transformation starts with rejecting the legacy finance systems architecture that provides only summary results. It is transforming the financial systems—the original enterprise data warehouse—into a system capable of supporting today’s information demands.
All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the author except for the use of brief quotations in a book review or scholarly journal. Posted by permission.
In 2004 IBM considered divestiture of GenevaERS to protect customer applications. This was the resulting memo summarizing operations and the business. More information about this time can be found in Balancing Act: A Practical Approach to Business Event Based Insights, Chapter 56. Walkabout
I EXECUTIVE SUMMARY
General Overview 2
Proposed Transaction 2
Reason for Sale 3
Key Reasons to Invest 3
II BUSINESS / INVESTMENT DESCRIPTION
Key Investment Highlights 5
Product features 6
Current Engagements 6
Overview of Geneva 7
GenevaERS Components 8
IV FINANCIAL SUMMARY
2003 Income Statement 10
Direct Costs 11
Direct SG&A 12
Balance Sheet Items 12
V MARKET OVERVIEW AND HISTORY 14
VI HUMAN RESOURCES 15
VII TRANSACTION SUMMARY
Transaction Description 16
Key Intellectual Property Terms for the Sale of GenevaERS 16
Partnership / Teaming Agreement 18
A Examples of Geneva Deployments 19
I EXECUTIVE SUMMARY
IBM desires to sell GenevaERS, the GenevaERS maintenance contracts and to give the potential Acquirer the option to solicit key code development employees. GenevaERS was formerly owned by PricewaterhouseCoopers Consulting and is presently a part of IBM Business Consulting Services (IBM BCS).
GenevaERS is a combination of a business intelligence application and a database query engine. It efficiently uses available memory, I/O devices, and computing capacity to resolve multiple business problems in a single pass through a database or data repository. It is a high-performance tool that enables data management and analysis procedures, which are typically impractical with standard tools. It is used in all aspects of the data warehousing process including extraction, transformation, loading, query processing, data mining and reporting.
IBM is seeking to sell GenevaERS assets, as outlined in Section III of this document, and the GenevaERS maintenance contracts to a Buyer who will continue to develop and maintain the tool on all of the existing deployments, including internal IBM installations. GenevaERS assets are to be sold “AS IS” and divested “WHERE IS”, subject in all instances to IBM’s ability to do so and consents of third parties.
Each potential Buyer will bear all costs associated with its own investigation of the acquisition and consequently with the transition.
In addition, as part of this transaction IBM BCS would consider a non-binding service partnership agreement with the Acquirer for Geneva deployment and marketing, should one be deemed appropriate, to facilitate future deployments of Geneva with adequate integration skills and expertise.
Reason for Sale
IBM has determined that Geneva, which it acquired as part of the PricewaterhouseCoopers Consulting acquisition, does not fit within IBM’s long-term strategy.
Because IBM customers have invested heavily on systems using Geneva, and IBM is committed to helping to protect that investment, IBM is seeking an Acquirer who would further develop the software and will maintain existing software deployments. The buyer is expected to specialize in business intelligence and data analytics software and ideally will have the following characteristics:
An acquirer will have significant size and strategic importance;
Ability to leverage the product across a wider client base;
An established software development organization in the Extraction, Transfomation and Loading (ETL) tools market segment which can further the capabilities, quality and flexibility of Geneva;
Ability and commitment to provide maintenance to existing customers for the duration of the contracts or for a period of no less than five years from the date of purchase which ever is greater.
II BUSINESS / INVESTMENT DESCRIPTION
Key Reasons to Invest
Acquisition of a unique and highly regarded set of tools
Geneva is one of few products that meet the growing business demand within a narrow niche for high-volume data analysis / ETL solutions. Its unique architecture allows it to outperform other competitive products. According to customer feedback, almost every Geneva implementation occurred after alternative approaches had been exhausted and alternative tools to solve the problem had failed. In these and other competitive situations, Geneva has demonstrated outstanding performance characteristics every time.
A set of tools that complements other product offerings
Geneva will enable a Buyer to extend the capabilities and the reach of existing products and solutions. Genevasales can be complementary to many existing products, and other tools can be complementary to Geneva. The more generic data tools are more flexible, yet inefficient and clumsy when they have to be used to solve specific problems. Geneva, on the other hand, is a highly specialized tool, which can be leveraged across a client base with massive data analysis processing power needs. Put differently, Geneva can complement a set of flexible generic tools, which have broader applicability but cannot to handle large volumes of data.
Access to several large clients
As part of the transaction IBM intends to transfer all Geneva maintenance contracts to the Acquirer. At present the GenevaERS Consulting Practice provides maintenance support to eight customers. Geneva has developed a loyal client base due to its unique capabilities. Customers have used Geneva on average for six years. A potential Acquirer can cross sell other data analysis products and services to any one of them.
Growing market opportunity
According to a recent report from Gartner the ETL tool market is growing due to the increase of enterprise size and correspondingly the growing importance of data integration in private and public sector organizations. Data volumes in every organization continue to grow, and now even medium-sized organizations routinely experience performance problems. If new marketing channels to distribute Geneva are established, significant opportunities may exist for new customers. In addition, many of the Geneva implementations differ significantly from each other in the type of problem solved. These successful implementations may present market opportunities in related areas.
Understanding of Key High Volume Principles
Geneva’s patented process uniquely combines many fundamental principles in solving high volume data integration and analysis problems. Geneva provides access to this knowledge, and many of these principles may be utilized in other tools and implementations.
Potential to solicit select key resources
Access to Geneva key developers is important for the delivery on maintenance contracts and for future enhancements of the software. Therefore, a potential Acquirer will have the opportunity to solicit key code development resources.
Alternatively, the GenevaERS practice can provide training to Acquirer’s existing organization and transitional services to existing customers. These arrangements will be a part of a separate set of contractual arrangements.
Key Investment Highlights
The value of Geneva includes:
Estimated revenue of $1.5M in 2003, of which $0.5 is from recurring maintenance contracts
History of profitability, estimated contribution margin of 42% in 2003
High growth potential
Description of the assets
The assets for sale include the following:
Version 4 source code written, in Visual Basic, C++, COBOL, MVS Assembler, PL/SQL, Java and SAP’s ABAP, including the following components
Workbench, allowing entry of Metadata and View Definitions
UNIX and MVS Performance Engines
Java and SAP Executive Information Report Viewers
Version 3 source code, written in COBOL, CICS, and MVS Assembler, including the following components
Metadata and View Definition Screens
MVS Performance Engine
Source code for various MVS Assembler and UNIX utilities which enable rapid Geneva API development, testing, debugging, etc.
Version 3 and Version 4 documentation
GenevaERS has been developed and marketed through consulting engagements. The small customer base reflects a sales approach of solving very difficult problems on typically very expensive consulting engagements lasting anywhere from six months to multiple years. Nonetheless, GenevaERS is at the core of a number of mission-critical processes for these clients, reflecting a sizeable investment in installation, configuration, enhancement and maintenance.
Cross selling of Geneva and other products and solutions to existing customer base
Leverage Geneva’s capabilities across a wider range of clients through aggressive marketing
Use Geneva’s mainframe capabilities as a complement to existing ETL offerings
The following are characteristics of many Geneva Implementations:
High-Volume data extraction and reporting requirements
Summary data structures are inadequate, requiring table scans
Large amounts of reference information and data reclassification
Batch window constraints or need to improve CPU utilization
Limited development time
Multiple Geneva implementations scan billions of records in tight nightly windows, resolving at times 1000s of queries, hardcopy reports, interface files, or report cubes in a single pass of the data. These processes often involve billions of joins to reference data. In some instances these join processes are date effective, providing query results as of a specific point in time. Implementations of Geneva have deferred hardware upgrades and made possible query processes on a daily rather than monthly basis. By being able to read from source systems, and do multiple tasks with the output in a single multi-purpose tool, systems can be built in a fraction of the time of other tools.
These statistics are not from artificially created benchmarks, but from mission-critical, real-world production systems at some of the most demanding customers with the highest volumes in the world. The tool has been engineered for performance on the specific platforms it runs on rather than ported from another.
One of the largest integrated electronic circuits manufacturers
One of the largest US food manufacturing companies
A large interstate bank
The Departments of Transportation of two States in Western United States
One of the largest integrated retail chains
Fortune 100 insurance company
The department of finance of a state in Northwest United States.
Overview of Geneva
GenevaERS provides information access and delivery capabilities with greater scale and performance than conventional methods. GenevaERS is a software product that addresses Business Intelligence areas including extraction, transformation, loading, query processing, data mining, and reporting. It uses available memory, I/O devices, and computing capacity to resolve multiple business problems simultaneously in a single pass through a database or data repository. Using parallel processing techniques and efficient algorithms, GenevaERS solves high-volume data analysis and reporting problems while maximizing the efficient use of computing resources.
The use of detailed transaction-level data as the backbone of Business Intelligence systems offers almost endless adaptability to changing business demands. However, the effort necessary to effectively utilize these high volumes should not be underestimated. Implementing workarounds that do not leverage the power of detailed transaction data can create larger, more pervasive, and complex problems in the long run. Our experience dictates that scale must be built-in; it cannot be added after the fact. Converting a one-story building to a five-story building is much more expensive than building a five-story building in the beginning. The same is true when working with Business Intelligence systems. GenevaERS applies proven methods and techniques to deal with the high-volume problems and scalability.
Geneva reporting and analysis systems can be divided into three different areas:
1) Data Sourcing or Extraction / Transformation / Loading (ETL),
2) Data Repository, and
3) Presentation Layer.
Typically, separate vendor tools are applied in each area. As indicated in the figure below, GenevaERS crosses these boundaries and has capabilities in each layer. It supports ETL, data access and manipulation, and presentation functions.
GenevaERS provides direct access to data stored in a variety of formats including sequential files, VSAM, DB2. Using Geneva’s multiple API points, it has been used to read IDMS, IMS, and many specialized or archaic formats. Similarly, GenevaERS can create several different output formats, including database ready load files, delimited, spreadsheet, columnar formatted hardcopy reports, and Geneva Executive Information cube format. The handling of multiple business functions in a single tool with the added capability of reading and producing different formats allows a user to solve difficult problems in a cost-effective manner.
GenevaERS is more than a combination of a Business Intelligence application and a database query tool. It can be used to build complex application subsystems usually created through custom development. Examples include multi-currency revaluation, billing systems, and cost allocations processes created by our clients.
Many tools have complex operating environments and user interfaces that require prior programming experience in order to utilize them proficiently. GenevaERS has a parameter-driven interface that eliminates the need for custom coding and reduces implementation timeframes. Additionally, the parameter-driven interface permits both non-technical and technical users to utilize the tool effectively. For example, approximately two hundred finance users create GenevaERS Views (i.e., queries) to produce reporting results at a state government client. At a manufacturing client, a handful of IT resources support the data repository and create load-ready data mart files using GenevaERS.
The Workbench is a parameter-driven front end that allows users to define metadata (i.e., input record layouts, physical file characteristics, and joins). Building GenevaERS Views or queries is handled by the Workbench through the GenevaERS View Wizard and View Editor. Users specify the input file(s), filtering logic, output layout and format, and processing logic required to meet business needs.
GenevaERS Performance Engine
The Performance Engine includes the programs and scripts required to execute GenevaERS Views. The Performance Engine consists of several steps required to efficiently process and format the input transaction and reference data.
GenevaERS Executive Information Report Viewer:
The GenevaERS Executive Report Viewer is a Java or SAP ABAP enabled application utlizing a high-performance proprietary access method and cube structure. It allows users to begin the process of understanding the results of a query at the high level, and drill down to the relevant details. It enables quickly changing dimensions to understand different points of view, export and download to spreadsheets, with all aspects of the cube produced by the performance engine in a fraction of the time of other tools. The Java component can be utilized as a web service, allowing customers to build their own presentation interface or on-line access of summarized results produced by Geneva.
IV FINANCIAL SUMMARY
Section IV contains 2003 results. The data has been prepared by the GenevaERS Consulting Practice management and it is for reference only as it relies on estimates extracted from the accounts of the practice and allocated to the GenevaERS software. All figures reflect a fiscal year-end of December 31.
Historical data is not audited by an independent auditor and additional details pertaining to historical data will be provided during due-diligence.
Geneva has traditionally been sold as a part of a service offering and thus service revenues were not separated from the license revenues. For the purposes of this confidential information memorandum the management of the GenevaERS practice has estimated the implied license revenues by taking a percentage of total services revenue that was considered attributable to GenevaERS Intellectual Property or product development. To determine the percentage of implied license revenue, client engagements were analyzed and the appropriate ratios derived.
Although IBM recently completed the latest version of the software to fulfill preexisting contract commitments to a specific client, it has chosen to stabilize the product, as at present it is not viewed as strategic. As a result, the revenues have grown marginally over the past few years. The potential acquisition of the Geneva software by a strategic buyer committed to Geneva’s continued enhancement can have a significant positive impact on the forecasted revenues. It is the understanding of Geneva management that customers are reluctant to commit a large investment to an infrastructure, which has continuously been viewed as an auxiliary tool of consulting services.
The direct costs are costs associated with the ongoing development and upgrade of Geneva. They include total compensation for code development, code maintenance and subcontractor expenses. All of the development and maintenance costs are expensed. Note that direct costs do not include general allocations.
Compensation is separated into development and maintenance components. The breakout for development and maintenance costs was determined by estimating time spent by specific resources on development, maintenance and consulting related tasks for the year. Development compensation includes the percentage of total compensation related to software development activities. Maintenance compensation is the percentage of total compensation associated with software maintenance activities. No bonuses or commissions have been paid historically.
Similar to compensation, subcontractor expenses are separated into development, and maintenance components. Development and maintenance of the software has involved key individual subcontractors. Two of these subcontractors have been involved in the tool from its inception, and a third for over five years.
Other Expenses include royalty payments to one of these subcontractors. In 1995 he agreed to exchange his intellectual property rights to Geneva for royalty payments on software license sales and services related to Geneva. Under the arrangement he is entitled to four and one-half percent of the first annual $15 million of license revenues and to one and a half percent of the first annual $10 million of Geneva related services revenues. These arrangements will expire when Geneva’s ownership changes.
Hardware costs related to hosting the GenevaERS software development and maintenance infrastructure may be incurred by the Acquirer.
Direct SG&A includes only marketing expenses attributable to the GenevaERS software, net of overhead allocations.
Balance Sheet Items
Geneva development costs were not capitalized, but expensed in each year as they arose. This was a result of the arrangements that the Geneva Practice had with customers who were billed against these expenses.
The prepaid maintenance of the contracts is estimated to amount to approximately $245,000 at the transaction closing date. The exact closing date may impact this amount.
All other assets or liabilities that will be transferred to a potential Acquirer are immaterial.
V MARKET OVERVIEW AND HISTORY
PwC consultants developed Geneva through numerous engagements, beginning in the mid 80’s. In 1993, Price Waterhouse began selling it as a commercial package. For many reasons Geneva did not become a larger commercial success icluding:
Geneva came to market just as the press was proclaiming the death of the mainframe, and Price Waterhouse’s revenue was driven by selling primarily non-mainframe solutions.
Geneva never enjoyed a dedicated sales force. Sales occurred as a result of informal referrals and at times consultants were reluctant to lose the appearance of being impartial by recommending a tool owned by their company, and not widely known in the trade press or by analysts.
Until the latest version was completed six months ago, Geneva had a character based user interface that was not appealing to end users.
PW did not have a software brand. Customers where reluctant to purchase software from auditors and consultants.
Sales were driven either by niche sales, or when the largest companies ran into problems that simply were too expensive to be solved by any other way if one was even possible.
Over time, the implementations of Geneva began to involve a Geneva “rapid response” of consultants who were very familiar with the tool, and would implement it in abbreviated time frames, at times under extremely high pressure to save a project or client relationship.
Given only a limited investment budget but having the ambition to tackle a very large set of problems, the Geneva team consistently strove to create a true software package. It created automated installation processes, maintained on-line access to customers, and provided customized training and on-site technical support. The team even manned a 24-hour help desk with consultants who would at times work with clients through the day and answer problems calls at night. They were able to periodically charge license fees ranging from $250,000 to $400,000 per “named application,” and in one case selling a million dollar site license.
However, because Geneva was embedded in a consulting organization, quite often Geneva was offered as a loss leader on proposals to obtain the consulting work. Routine maintenance was funded by the annual maintenance contracts. However, new development was primarily funded though specific engagements where a specific enhancement was required.
Due to the complexities of the projects and the required enhancements, what began as primarily a mainframe reporting tool has become an attractive combination of a tool capable of significant ETL processes with certain reporting capabilities. Working with the premier PwC client base of the largest organizations in the world kept the tool focused on performance.
The result is a tool:
That has developed outside the constraints and “conventional wisdom” of traditional software tools;
That cuts across traditional software categories;
Has innovative features developed at times for one client in response to real-world problems, and
Performs at the highest levels.
The average Geneva client has used the tool for approximately six years despite continued efforts by other companies to persuade clients to implement more ‘mainstream’ ETL products. For example, one client has expended a significant amount of resources over four years to implement a Teradata solution, and the implementation is still more than two years away to replace a Geneva solution which was implemented in about 18 months.
PricewaterhouseCoopers made limited attempts to market Geneva. Because of auditor independence requirements, this dictated a very narrow set of potential buyers: Only companies which were not audited by PwC or substantially owned by someone audited by PwC. Thus the potential customer base appeared to be largely constricted by requirements unrelated to Geneva.
IBM acquisition of PwC Consulting eliminated these restrictions. However, Geneva does not fit with IBM’s strategic data management software direction of focusing on the database, and raises problems of database independence for an ETL tool. Besides, the continuing residence within the consulting practice does not eliminate many of the marketing issues Geneva has had for many years, and raises doubts with customers about the viability of the product.
With the completion of the latest version, including late stage beta testing by the largest customer, the software is substantially ready for the market. Preliminary marketing efforts have been favorable. With the appropriate marketing, possible integration with established tools and the assurance of long-term support, Geneva is likely to have significant potential in the coming years.
VI HUMAN RESOURCES
As part of this transaction, the GenevaERS Consulting practice, is offering the option to potential Acquirers to solicit select key development staff in order to facilitate the uninterrupted service on existing maintenance contracts.
To date Geneva has been developed and implemented by a team of consultants. Thus, most of GenevaERS staff are flexible and have significant customer interaction experience.
The select key development personnel have extensive code development experience in algorithmic and machine-level computer languages. They have been at the heart of development efforts and are intimately familiar with Geneva, its structure, modules and algorithms. All of them have at least 10 years of software development and consulting experience.
VII TRANSACTION SUMMARY
IBM BCS would like to divest the GenevaERS software assets to an Acquirer with strong development capabilities in business intelligence and data analytics tools who will continue to develop, enhance and maintain Geneva. The Acquirer is expected to specialize in business intelligence and data analytics software and ideally will have the following characteristics:
An acquirer will have significant size and strategic importance;
Ability to leverage the product across a wider client base;
An established software development organization in the Extraction, Transfomation and Loading (ETL) tools market segment which can further the capabilities, quality and flexibility of Geneva;
Ability and commitment to provide maintenance to existing customers for the duration of the contracts
The assets to be divested include:
Version 4 source code
Version 3 source code
Source code for various MVS Assembler utilities which enable rapid Geneva API development
Version 3 and Version 4 documentation
Eight maintenance contracts with high-powered customers
An option to solicit select key employees
Key Intellectual Property Terms for the Sale of GenevaERS
The purpose of this outline is to summarize key intellectual property terms under which IBM proposes to sell certain assets related to GenevaERS. This outline is meant to serve as a general framework. Further discussions may alter the terms outlined below or add new terms prior to execution of definitive agreements.
1. Sale of Software and Documentation.
Subject to certain preexisting obligations, IBM will assign to Buyer its rights, title and interest in the copyrights to the source code, object code, and related documentation of the GenevaERS product (collectively referred to as “GenevaERS” or “Geneva”).
2. Reservation of Rights
IBM will retain the right to use code fragments of GenevaERS, where code fragments are defined to be not more than 30% of the total lines of GenevaERS code, to support IBM’s products and services. IBM will retain the right to use GenevaERS for existing internal GenevaERS installations and for fullfilling any existing contractual obligations.
3. Product Name
Subject to any preexisting rights which may exist with third parties, IBM will quitclaim any rights it may have in the GenevaERS name to Buyer.
4. Conditions on Sale.
GenevaERS will be assigned “AS IS”, “WHERE-IS” without warranty, including, without limitation, any implied or express warranty of fitness for a particular purpose or of non-infringement of third party intellectual property rights. IBM will not provide support for GenevaERS.
The Buyer will be responsible for obtaining any other software and licenses upon which GenevaERS depends from the applicable vendors.
6. Governing Law.
The agreements contemplated hereby will be governed by the substantive laws of New York State, without regard for its conflict of laws provisions.
7. Freedom of Action.
Any definitive agreement shall provide that, subject to the other party’s intellectual property rights, there are no restrictions on either party’s ability to (i) provide any goods, services, or intellectual property to any other entity or (ii) form any alliances or take an equity or other interest in any other entity.
The Buyer will defend, indemnify and hold harmless IBM, its subsidiaries, and its and their directors, officers, agents, employees and representatives from any claims, causes of action, and expenses, including reasonable attorneys’ fees, arising out of or related to the Buyer’s use of GenevaERS.
9. Exclusion of Damages.
Neither party shall be responsible for any indirect, special, punitive or consequential damages, whatsoever, including loss of profits or goodwill, in connection with any aspect of this transaction.
Any press releases, announcements or other forms of publicity concerning the transaction contemplated hereby will be coordinated and approved in writing by both IBM and the Buyer.
The Buyer will pay IBM a one-time fee at agreement execution in consideration for GenevaERS.
Partnership / Teaming Agreement
In addition to the outlined terms, as part of this transaction IBM BCS would consider a non-binding service partnership agreement with the Acquirer for GenevaERS deployment and marketing, should one be deemed appropriate, to facilitate future deployments of Geneva with adequate integration skills and expertise.
Geneva Enables All Phases of Data Warehousing from
[The document then contains samples of citations listed in other documents.]
 Historical data prior to IBM’s acquisition of PWCC was prepared under different internal accounting standards and thus represents a neither comparable nor meaningful data reference. Furthermore, in 2002 the GenevaERS practice was engaged primarily on PWCC internal projects.
 The estimated license revenue line includes revenues from internal engagements, approximately 28%. On internal engagements, the GenevaERS practice was reimbursed at cost.
 Neither bonuses nor commissions have been paid historically.
 Direct SG&A does not include overhead allocations.
The following are two solution blueprints by an additional attempted start-up to license Geneva. In this instance, the platform was to be ported to the Intel Itanium processor, putting a very powerful data analysis tool right under the desk of the analyst.
The vision was far sighted, anticipating what happened with cloud and data analytics. But it was ravaged by the technology collapse post Internet bubble.
INTELLIGENCE ENGINE FROM PwC CONSULTING AND DIGINEER™
Consolidating massive amounts of data from multiple sources to gain an integrated view
OI5 DataEngine™ from Digineer™
Enterprise Hardware Platform
Scaleable Intel® Architecture servers
Advanced Transaction Analysis for Business and Government
Crucial data that can help businesses and government operate more efficiently often lies buried within an overwhelming mass of transaction details scattered among numerous computers. Extracting this data using most data mining solutions presents many difficulties, since these solutions lack the flexibility, scope, and sheer processing power necessary to convert uncorrelated data into useful information. The Intelligence Engine solution combines technology expertise from PricewaterhouseCoopers (PwC Consulting), Digineer™, and Intel to help solve this issue. The Intelligence Engine solution spans data repositories across the world, using query-processing techniques to generate data intelligence from large numbers of events and transactions.
FACING THE CHALLENGE
Without the ability to look deep into operational processes, large-scale businesses are hindered in their decision-making and organizational goals––essentially running blind. Where data exists, it often resides on disparate servers in a variety of formats, resisting easy access or useful analysis. Collecting, consolidating, and correlating this data requires tools that can be scaled for massive volumes of data, and adapted easily to expose a wide range of information. Challenges to this end include:
Huge volumes of transaction data––Many conventional data warehousing and data mining solutions collapse under the weight of millions of daily event-level transactions. A robust solution must marry intelligent query processing with sufficient processing power to meet high levels of activity.
Data scattered throughout many repositories––Large organizations frequently store data in a variety of formats in data repositories spread throughout the country––or the world. Finding the key data—the proverbial needle in a haystack—demands a solution scaled to the magnitude of the task.
Need to quickly spot trends and patterns––The accelerating pace of modern business makes it necessary to detect and respond to trends and patterns in days rather than weeks. Knowledge-discovery tools must be flexible and adaptable for this purpose.
Evolving data requirements––Creating data models and report formats can be a very time-consuming and expensive proposition, and updating them can also require excessive time and effort. Tools to reduce this design effort and the associated maintenance help ensure that changing data requirements can be accommodated in minimal time.
MEETING THE CHALLENGE
Being able to accurately detect emerging patterns and trends in daily operations is invaluable to any organization seeking to remain competitive under the pressures of rapid change. The Intelligence Engine solution incorporates Digineer™ OI5 DataEngine™ deployed on Intel® Itanium® processor family-based servers to provide fast and flexible analytical data handling across multiple systems. By providing consolidated views of dispersed data in a variety of formats, this solution lets large organizations make fully informed decisions within their industry by assimilating data from every group or department––regardless of platform. The OI5 DataEngine™ successfully bridges dissimilar data structures and geographically separated servers to turn raw data into useful knowledge, providing a robust solution to help guide leaders of large organizations.
To accelerate the process of collecting and consolidating data, Digineer’s toolset simplifies query-processing setup and report generation. PricewaterhouseCoopers has successfully deployed this solution in a wide range of industries, including retail, financial, healthcare, and government.
The Intelligence Engine solution lets any large organization benefit from improved visibility into the heavy volume of daily transactions and events specific to their business––whether retail sales tallies per product for a nationwide distributor, the weather conditions compared to yields at a juice producer’s orange orchards, or quality control data tracked by a steel manufacturer looking for ways to improve processes.
Clients adopting this solution enjoy improved operations management through rapid analysis of transaction data. They also gain deeper insights into key processes within their organization that may have been inaccessible in the past because of the difficulty in collecting and correlating the relevant data.
FEATURES : BENEFITS
Supports multiple data formats: The OI5 Data Engine™ provides support for a wide range of data formats, allowing flexible handling of data stored in repositories across many different platforms.
Designed for faster, high-volume transaction processing: The Intelligence Engine solution combines the intelligent pre-processing capabilities of the OI5 DataEngine™ with the architecture benefits of the Intel® Itanium® processor family. The result is a solution that scales well for the extremely large volumes of data involved in transaction processing––producing rapid responses to complex data queries.
Adapts easily to changing data requirements: The solution includes tools that simplify query design and report generation, encouraging companies to create new models for data collection and analysis, while meeting the core data extraction requirements for the organization.
Advanced Business Intelligence and Analytics Solution
Retailers, financial institutions, healthcare companies, and government agencies are collecting more data than ever before. This includes sales data from stores, warehouses, e-Commerce outlets, and catalogs, as well as clickstream data from Web sites. In addition, there is data gathered from back office sources––merchandising, buying, accounting, marketing, order processing, and much more. To pull together information from all these different sources and integrate, process, and analyze it, companies require advanced business intelligence and analytical solutions powered by a robust, high-performance infrastructure.
The Intelligence Engine solution addresses this very need, by combining technological expertise from PricewaterhouseCoopers (PwC Consulting), Digineer™, and Intel®. The query-processing capabilities of the Digineer™ OI5 DataEngine™ combined with the processing power of the Intel® Itanium® processor family, enables PwC Consulting to provide companies with the ability to quickly consolidate and integrate massive amounts of data from multiple sources––giving key decision makers a single view of vital business information.
Optimized to run on Intel® Itanium® processor family-based platforms, the Digineer™ OI5 DataEngine™ delivers a reliable, high-performance mechanism for processing and querying vast quantities of raw, event-level data through a software algorithm that employs intelligent pre-processing for improved efficiency. The Explicitly Parallel Instruction Computing (EPIC) technology and 64-bit addressing capabilities found in Intel® Itanium® architecture, deliver new levels of performance when applied to data mining and knowledge discovery operations. Powerful Intel® Itanium® processor familybased servers provide the massive computing power needed to run and consolidate memory-intensive databases across channels. In addition, integrating different processes and organizations is easier and more cost effective with a modular, open environment based on industry standards and Intel® Architecture.
THE BUSINESS CHALLENGE
Today’s competitive advantage will go to companies that make the right decisions at the right time. To succeed, these companies must be able to quickly consolidate and integrate massive amounts of data from multiple sources, into a single view of their business that can increase marketing and operational efficiency.
Turning Data into Dollars, a May 2001 Forrester report, indicates that executives charged with setting the strategic direction of their organization understand the value of business intelligence and analytics, yet are faced with the following challenges:
Mountains of data create processing bottlenecks—The sheer volume of data collected–– often involving terabyte-scale databases––creates the potential for processing bottlenecks in those systems not designed to effectively cope with this quantity of information. The massive data quantities produce such a heavy processing load that many existing solutions rapidly reach throughput limitations. Without sufficient processing power and highly optimized analytical tools, organizations must often relinquish access to important subsets of their data, or must add expensive hardware and software to increase system performance.
Analyzing information from scores of data repositories presents a challenge— Organizations cannot easily access and analyze their operational information because of the difficulty in extracting large quantities of data from multiple data repositories spread throughout the organization. Existing solutions lack the flexibility and robustness for effective data access and extraction.
Inflexible models fail to support evolving data streams—Organizations spend significant amounts of development time creating data models, collecting data, defining reports, and constructing consolidated data warehouses. In highly dynamic environments, these systems may rapidly lose relevancy and require ongoing adjustments at the core software or hardware level. Updating these core data model components to correspond with evolving knowledge discovery requirements necessitates a significant investment in time and effort.
New analytical views must be developed quickly and on demand—Different departments and groups within an organization have unique needs for the information mined from available data, and these needs change frequently, often on an urgent basis. A practical solution must have the capability of easily providing new analytical views from the data to respond to immediate needs from diverse groups.
Extracting key information is inordinately difficult—The ability to drill down through individual transactions and extract the useful patterns and trends is critical to knowledge discovery and analysis. Most existing systems do not adequately provide this capability when there are large quantities of data stored on multiple systems, or in one large system where data access is cumbersome.
The retail environment faces several challenges that can take advantage of a business intelligence solution like Intelligence Engine. For example, in a typical retail environment, data gathered from stores, Web sites, catalogs, as well as suppliers and warehouses provide valuable kernels of knowledge that can increase profitability. Consolidating and analyzing massive amounts of memory-intensive databases requires a robust infrastructure. As a result, many retailers mine only a subset of their available information. To recognize buying patterns over time, massive volumes of sales transactions must be analyzed and correlated by sales channel, region, individual store, product, and customer preference. Within the typical retail environment, this data resides in a number of separate systems.
THE SOLUTION OVERVIEW
The Intel® Architecture-based Intelligence Engine solution provides companies with a fast, affordable, and flexible way to consolidate, integrate, and analyze massive, memoryintensive data across multiple systems.
The Digineer™ OI5 DataEngine™, operating on a platform powered by the Intel® Itanium® processor family, performs high-speed analytical processing of massive amounts of event-level data. The data engine can simultaneously write large numbers of data extracts to downstream data warehouses, online and offline reporting tools, and real-time analytics applications. The 64-bit addressing capabilities of the Intel® Itanium® architecture provides the robust computing power needed to run and consolidate memory-intensive databases across channels. The extraordinary performance of the Intel® Itanium® processor family, coupled with the EPIC features, provide an architecture fully capable of adding capacity as needed, and providing heightened performance.
Historically, retailers have addressed the scale and performance problem by creating large data stores using conventional technology. As the volume of data that must be captured and processed increases––often exponentially––processing the massive volumes of data creates a bottleneck and prevents timely access to the detailed information required to run the business. Once this critical point has been reached, retailers must choose between adding expensive hardware to boost performance, or sacrificing the breadth and depth of captured data.
The OI5 DataEngine™ offers an attractive alternative to the scale and performance problem. Retailers use the solution to analyze key retail metrics, such as customer profitability, recency, frequency, volume analytics, supplier performance, and enterprise financials. The OI5 DataEngine™ derives these analytics from a collection of customer relationship management (CRM), supply chain management (SCM), and point-of-sale (POS) databases–– even if the data is distributed among several different machines. The performance provided in these kinds of implementations costs much less than equivalent systems.
The Digineer™ OI5 DataEngine™ delivers:
Speed—Performs fast and efficient processing of massive volumes of transactional data, resulting in a consolidated data store that accurately represents the organization’s operational dynamics.
Scale—Integrates data from multiple, disparate silos without requiring any modifications or enhancements to the existing data sources.
Depth—Includes powerful data query capabilities that efficiently and economically enable data mining of massive amounts of transaction-level event data.
Breadth—Provides an outstanding price and performance ratio, making the solution accessible to medium- and large-sized organizations.
PwC Consulting fully integrates the OI5 DataEngine™ into the client’s business processes and technologies. PwC Consulting’s background and experience in business intelligence and analytics produces a solution that enables organizations to rapidly collect, aggregate, manage, analyze, filter, customize, and distribute information throughout their value chains. By providing consolidated views of the dispersed data––rapidly and cost-effectively–– organizations can identify trends and quickly respond to changing needs and situations, thus improving overall business performance.
The Intelligence Engine solution runs efficiently on a 4-way Intel® Itanium® processorbased server. Supported operating systems include (Win64) Microsoft* Windows* 2000 server (Q3’02), Linux* (Red Hat*, version 7.2 for the Itanium® processor), and HP-UX* version 11 (Q4’02).
Components of the Digineer™ OI5 DataEngine™ rely on the following technologies:
Core Data Engine™—C++ with multi-threading
Result Set Viewer—Java* using Swing*
MetaData Manager—Visual Basic*
MetaData Database—Any ODBC-compliant database
The OI5 DataEngine™ runs optimally using Intel® EPIC architecture, taking maximum advantage of the Itanium® processor family’s multiple memory registers, and massive on-chip resources. The 128-bit general and floating-point registers excel at supporting the complex operations involved in analytics. Tuning the implementation to achieve maximum performance through parallelism, and tapping into the power of the Itanium® processor family’s multiple execution units makes it possible to rapidly process and analyze the large volumes of data necessary to accomplish this solution.
PwC Consulting solutions employ a variety of technologies using Digineer™ OI5 DataEngine™. By being able to create data stores through the importing and exporting of data files, this solution makes it unnecessary to devise complex, custom integration schemes. Support for each of the following sources and targets is provided:
Microsoft* SQL Server*
Target analysis engines
EMC* Symmetrix Manager*
IBM* Intelligent Miner*
Cognos* Powerplay* and Improptu*
Custom applications built in Visual Basic, Java, C++, and HTML using Microsoft Internet Information Server* (IIS) and Transaction servers
Digineer™ OI5 DataEngine™ has been deployed for use with SAP* and PeopleSoft* ERP packages as sources for data and targets for additional analysis. Future plans include development of interfaces for MQSeries*, and clickstream data sources, scheduled for release by mid-2003.
WHO THE SOLUTION WILL BENEFIT
The Intelligence Engine solution provides rapid analysis of data from multiple sources to gain a better picture of emerging trends and patterns. For a strong competitive advantage, companies will need to develop a quantitative, as well as qualitative, understanding of their business. The Intelligence Engine solution complements existing data warehouse and business intelligence solutions. It can also maximize the performance and extend the functionality of systems already in place.
The key vertical integration areas in which the Intelligence Engine solution can provide significant value include:
Retail—Domestic and international (Western European) retailers with annual sales of more than $1 billion, and large-scale customer, supplier, point-of-sale, and operational databases.
Manufacturers—Mid-market manufacturers with annual sales of more than $500 million and one or more large silo databases.
Financial Services—Financial services organizations with multiple geographic locations and stringent operational and regulatory reporting requirements.
Healthcare—Pharmaceutical and biotechnology companies that depend on data from multiple systems to manage and optimize their operations.
Government—Agencies under pressure to provide more extensive reporting of data, over which the organization has little governance. In addition, data sources continue to propagate, grow in size, and are increasingly complex and divergent from each other. 5
The PwC Consulting Intelligence Engine solution using Digineer™ OI5 DataEngine™ adapts well to environments where operational data is distributed among several different systems––sometimes in different data formats––and where the report requirements change frequently. Implementing the Digineer™ OI5 DataEngine™ on servers powered by the Intel® Itanium® processor family offers these benefits:
Provides faster high-volume transaction processing. Supplies the massive computing power needed to run and consolidate memory-intensive databases across channels. This enables key decision makers and business owners to obtain accurate business intelligence and analytics, and quickly gain deeper insights into the dynamics of their operations.
Supports multiple data formats. Supports multiple applications and loads required to consolidate and analyze data throughout the company. This enables the solution to easily access existing data stores and business intelligence solutions without modifications or retrofitting.
Adapts easily to changing data needs. Accommodates an organization’s evolving data needs and models so that new informational requirements can be met with minimal investment of time and money. Integrates different processes and organizes these in an easier and more cost-effective manner with a modular, open environment based on industry standards and Intel® Architecture. Solutions based on the OI5 DataEngine™ as deployed on servers powered by the Intel® Itanium® processor family provide a highly competitive price and performance ratio. 6
FUNCTIONAL BUSINESS CONCEPT
Maintaining a current and accurate view of an organization’s operational status requires the capability of locating and analyzing information from a very large data store in a short period of time. This problem can be compounded when data resides in disconnected data silos, or has been extracted from a variety of non-correlated reports. Reducing the scope of the data analyzed in order to save time or expedite processing can create an inaccurate view of current operations. Processing bottlenecks and inefficiencies often result from solutions implemented on platforms without a tuned architecture capable of handling data quantities that reach terabyte levels. While increasing the quantity and complexity of a solution’s hardware and software may improve performance, it also raises costs. A better approach is to deploy an implementation on a platform designed to accommodate massive volumes of data with processing power that is suitable to the task. The Intelligence Engine solution using the OI5 DataEngine™ on a platform powered by the Intel® Itanium® processor family performs very high-speed, analytical processing of massive amounts of event data. This event data can be dispersed across any number of data sources, including enterprise resource planning (ERP), SCM, CRM, and legacy systems–– without reducing the effectiveness of the solution. Processed data can then feed downstream data repositories, analytical and business intelligence tools, and executive reporting systems, as shown in the following diagram.
The OI5 DataEngine™ uses a patented software approach that leverages the latest advances in microprocessor technologies featured in the Intel® Itanium® architecture, including 64-bit addressing and parallel processing capabilities.
The open architecture of the OI5 solution allows data to be extracted and transformed directly from legacy systems. OI5 provides a massively scaleable, high-performance solution for the most demanding intelligence and analytics applications
The first step to using OI5 is to define the metadata, including record layouts, files or tables, and relationships between tables. The OI5 design removes the complexities of understanding table relationships. Common keys and other relationships between tables only need to be defined once. Instead of having to understand the “where” statement in SQL to add department names to a sales report, OI5 lets users select the department name and then add it to their report. OI5 automatically locates the correct data.
Next, define the views (queries):
Select the transaction files to read
Specify the appropriate output format
Indicate the appropriate filters to apply
Indicate the data to sort or summarize
Design columns, and indicate any calculations to perform
OI5 does the rest
During data file output, OI5 produces multiple files in one execution of the process. This feature reduces the time needed to supply data to other systems. OI5 views can reformat data,perform logic operations, and make calculations. With the open API, custom functions can be created for other types of transformation. These features make OI5 a powerful data extraction and transformation tool, as well as a flexible application engine.
The OI5 DataEngine™ helps large organizations meet the scaleability and performance challenge of assessing key organizational operations when the number of events being measured and analyzed becomes too great. The solution supports queries directed to extremely large volumes of event-level data in a single pass.
Data exposed by the OI5 DataEngine™ can reside in different physical locations on multiple systems, including ERP, CRM, SCM, and POS systems. The front end of the OI5 DataEngine™ maps these “in-place” data sources into the data engine’s MetaData definition modules. The engine then uses the MetaData definition to intelligently pre-process the queries before generating the data views and outputting data stores for analytical processing. The data output can either be viewed with the OI5 Viewer or directed to other business intelligence and reporting systems.
The core, patented technology of OI5 is configured for refreshing and managing a largescale operational data store (ODS) derived from multiple data sources. As shown in the following diagram, it includes three integrated modules that deliver the full power of the OI5 technology and its massive data throughput: OI5 MetaData Manager, OI5 View Manager, and OI5 Core Processor. The OI5 DataEngine™, as implemented on servers powered by the Intel® Itanium® processor, runs on Microsoft* Windows* 2000 Server (64-bit) and HP-UX systems.
The parallel query-processing capabilities of the system follow a defined software algorithm that is optimized for speed. The algorithm first pre-processes all the data queries based on the data engine’s knowledge of where and how the event data is structured and formatted (determined by the MetaData definition in combination with Read Exits), as well as the required output data stores (known as Views). Optimized code generated by this query pre-processing step gets sequenced into a logic table. When the logic table is processed, the OI5 DataEngine™ spawns parallel query instructions that minimize instruction path lengths to the available microprocessors. This explicitly parallel query-processing directly takes advantage of the EPIC architecture of the Itanium® processor family. The OI5 DataEngine™, using instruction path optimization and single-pass data queries, offers unmatched query-processing and exceptional handling of the data view output.
The OI5 DataEngine™––the basis of this data warehouse analytics solution––is optimized for the Intel® Itanium® processor family. The following diagram shows how information from existing data warehouses is delivered to the OI5 DataEngine™ for advanced analytics processing. The OI5 MetaData and View Manager applications, running on a workstation powered by the Intel® Pentium® 4 processor, define the relations and views for the associated processing and analytics. Processed data can be directed to Web portal solutions, as well as custom applications and reports, or supplied in a form for re-entry into an existing data warehouse.
The Web portal is powered by the Intel® XeonTM processor family. The OI5 DataEngine™ is powered by a 4-way Intel® Itanium® processor-family based server.
PwC Consulting has used the OI5 DataEngine™ in numerous scenarios that have been designed and deployed for clients. This solution achieves favorable results in knowledge discovery applications that involve massive numbers of transactions. Running on servers powered by the Intel® Itanium® processor family, the underlying OI5 software data engine performs very high-speed, batch-analytical processing of large volumes of event data–– while simultaneously writing multiple data views to downstream data warehouses, online and offline reporting tools, and real-time analytical processing applications. When used in combination with Intel® Itanium® processor family-based platforms, the OI5 DataEngine™ offers an exceptional price and performance value.
*Other names and brands may be claimed as the property of others. Information regarding third party products is provided solely for educational purposes. Intel is not responsible for the performance or support of third party products and does not make any representations or warranties whatsoever regarding quality, reliability, functionality, or compatibility of these devices or products.
An R&D statistical study at a Fortune 100 airline suggested that millions of dollars each month were being lost due to lost revenue. The cause of the lost revenue ranged from simple ticketing mistakes and training issues to intentional fraud. But the approach used in the R&D study had the following limitations:
Statistical sampling does not identify specific cases. To take action specific tickets must be identified. Also, some tests require identifying patterns for individual employees or customers, which requires very large samples, preferably the entire database.
Sources other than the actual ticketing database were used because of its complexity and critical production system availability. But working against anything less than the production ticketing database introduces the possibility that the results can be disputed.
Not all tickets could be inspected because of the volume of ticketing data. Over 500 million records, from approximately 40 entities comprising 40 million tickets needed to be scanned. This had to be done for over 10 different specific lost revenue detection tests.
Certain types of detection tests could not even be attempted in the study because of the complexity of the logic involved.
The airline agreed to put Geneva ERS to the test. In a 14-week effort, a nine-member team performed the following:
The detection test business logic was defined, and data mapping from the business logic to the database performed,
Custom code was developed to scan the CA IDMS ticket database, and execute other complex logic
An architecture was developed, Geneva ERS installed, and the database structures defined within the tool,
Geneva ERS “views” or queries were created to produce four files (virtual and physical) and over 10 violation reports,
The queries were executed and refined dozens of times against test databases about 1/6th the size of the production database.
Executions against the production database required scanning the 500 million records in approximately 1 ½ to 3 hours wall clock time and from 3 to 6 hours CPU time. The ticket database was scanned using 30 parallel processes, ultimately reading 170 different files. All detection tests were resolved in one scan of the production database.
The results validated the dollar values estimated in the R&D study showing that over $6 million annually were being lost in one area alone, and millions more might be reported incorrectly. It also provided insight into some areas that had never been investigated before. But more importantly, Geneva ERS identified specific cases which could be investigated and collected. The evidence was so solid that certain employees were dismissed as a result of the investigation.
The following deck proposed similar projects to other airlines.
[An investment bank arm], a division of [a major Swiss bank], chose Geneva ERS to manage its Detail Repository, a central component of the Global General Ledger SAP implementation. This repository provides trade-level detail analysis to for internal and external global reporting. [Bank] management has referred to the global general ledger as “a cornerstone of [the bank]’s strategy and its successful realization globally is vital to all future plans.”
Approximately 1.3 million detailed transaction data records from twenty-two different feeder systems are loaded into the Detail Repository nightly. These transactions are trade level detail records from Europe, Asia Pacific, and North America. Geneva ERS scans the repositories’ 51 million records in 22 entities and 269 physical partitions. It extracts 20 million records that are aggregated into approximately 480,000 summary balances. These summary records are sent to SAP for balance sheet and summary profit and loss reporting. This process runs in approximately 3 hours of elapsed time and 5 and ½ hours of CPU time and produces 30 different outputs.
A second Detail Repository process uses Geneva ERS and custom programs to satisfy the intricate regulatory requirements. This system consists of 65 Geneva “Views” or programs, 4 custom programs, and 5 PERL scripts. Geneva is executed 19 times with each execution handling a subset of the core business requirements. During this nightly process Geneva reads 71 million records in 40 gigabytes, extract 59 million records in 30 gigabytes, and performs 229 million table joins. The output is created in 12 CPU hours and 8 wall clock hours. In comparison, legacy applications required 24 hours to complete a limited number of these processes.
Outputs from these processes are used for US tax and regulatory, Asia specific regulatory management, and Swiss regulatory reporting. They include information on:
Average Balancing and Multi-Currency Revaluation
Syndicated Loan Netting
Federal and Swiss Regulatory Collateral Allocation
Interest Rate Selection
Product Risk Weighting
Specific Reserve Allocation
Unutilized Facility Calculation.
The view outputs include files used in additional processing or to feed other systems, delimited files, tabular reports, and inputs to a sophisticated executive information system. The executive information system allows users to select which report to view and for what period. The user is presented with the highest level summaries. The user can then drill down into specific areas of interest, select ranges of data, sort columns, view other dimensions of the same data, graph the data, and export to spreadsheets. The executive information system is accessed by as many as 50 users throughout the world.
The Geneva Views are maintained in Sybase tables accessed by Geneva ERS Visual Basic ViewBuilder front-end. The software maintains various types of metadata including record layouts, field types, join relationships between record structures, logical and physical file partition information, as well as the actual query selection, summarization, and formatting logic. The business logic contained within the views ranges from simple transform logic to the sophisticated globally defined business rules that make up the global general ledger accounting model.
Insurance Company Manages Very Large Data Warehouse
Three years ago, a Fortune 100 Insurance Company embarked upon a “ Data Warehouse Program” to “build a single consistent common source of financially balanced data to be used by multiple business users for decision support, data mining, and internal/external reporting.” A major component of this information architecture is the construction of multiple Operational Data Stores or ODS’s. These ODS’s may contain up to 10 years of event level or detailed transaction history as the basis for populating other data warehousing applications. The company uses Geneva ERS to load and extract data from these data stores.
Data is loaded into these repositories nightly. In some cases Geneva ERS is used to extract data from operational source systems. “Source systems” vary from the PeopleSoft Journal Lines DB/2 table to legacy flat files. The DB/2 extract process reads 80 physical DB2 partitions in parallel, scanning over 80 million records in approximately 10 minutes.
This extracted data is edited using a combination of custom Geneva ERS processes and stand-alone custom programs. After the data is edited it is loaded into the repositories in parallel by Geneva ERS. The number of records loaded varies by repository, from 100,000 financial transactions (journal lines) to more than 1,000,000 policy transactions. In total, approximately 2 million transactions a day are added to the repositories.
In the same pass of the data that loads the repositories, Geneva ERS also produces multiple reports and extract files for use by business users and other application systems. One of the repositories is “backed up” by Geneva ERS at the same time. This daily execution of Geneva ERS reads over 450 million records in 23 different entities and 220 physical partitions. It writes over 900 million physical records to over 800 files. This process has run in 1 hour 3 minutes of wall clock time and 3 hours of CPU time.
Other executions of Geneva ERS extract data on a daily, weekly, monthly and annual basis. The output from one execution creates an executive information drill down file accessed via the company Intranet. This web site is accessible by over 300 business users.
The Geneva “Views” executed in all of the above processes are maintained within the Geneva ViewBuilder software. This includes record structures, field types, relationships between structures, and the actual queries themselves. Many queries are very similar to programs in their sophistication and contain complex business logic. Some have over 900 column definitions. These views also utilize custom code for accessing certain files and executing business logic requiring a programming language. Over 100 people have been trained on programming using Geneva ERS, and the company has had up to 10 testing environments at one time.
The most sophisticated use of Geneva ERS emulates a PeopleSoft financial cost allocation process. Custom programs were developed which generate over 6,000 Geneva ERS views based upon over 7,000 PeopleSoft rules. Geneva ERS executes these views to scan the financial repository selecting records eligible for allocation. It then allocates these costs through four allocation layers, such as products, and geographical units. At 1999 year-end, this process read over 50 million records selecting nearly 3 million that were eligible for allocation. These 3 million records were exploded into over 290 million virtual allocation records, of which 186 million summarized records were written to physical files. The process runs in 7½ hours wall clock time and 28½ hours of CPU time.
Financial Services Company Simplifies Complex Processes
Many financial service organizations were early adopters of computer technology. They quickly constructed systems to automate numerous processes. Programs were often added to the fringe of existing processes to solve new problems or provide an additional report or file for analysis. The need to keep multiple types of books and fulfill regulatory reporting requirements added to the complexity of some of these systems. These systems grew before modularization, subsystem definition, or other computing concepts were developed. Furthermore, the number of transactions generated by these organizations always challenged computing capacity. This required creative and complex solutions. Over time these systems metamorphosed into closely intertwined, interconnected, and inflexible systems.
In 1996, a Fortune 100 Insurance Company determined their reporting system was so inflexible that it would become unsupportable in the near future. They decided “…to build a single, consistent source of financially balanced data to be used by multiple business users for decision support, data mining, and internal/external reporting.” After analyzing their information needs they determined that the best approach to satisfy the broadest number of users was to construct a comprehensive data warehouse environment, including extract transformation processes, Operational Data Stores (ODSs), reporting environments, and data marts. The company viewed the ODSs as the heart of the data warehousing effort. ODSs are designed to contain up to 10 years of detailed transactional history. By storing the transactional detail, the company can satisfy requests for individual transactions, summaries using any field from the transaction detail, or snap shots as of any point in time. This robust environment truly enables the “single, consistent source of financially balanced data.”
Although such a data store simplifies and satisfies all types of information needs, it also means tremendous data volumes. Managing such data volumes requires a creative, industrial strength solution. It also requires people who know how to make it happen. This company went looking for a tool and the people that were up to the challenge. They chose PricewaterhouseCoopers and Geneva ERS.
Geneva offers a robust data warehousing solution able to process tremendous amounts of data quickly. It also provides for innovative, flexible solutions that are easily supported by company employees. This company uses Geneva to do the following:
Source system Extract and Transform
ODS Load, Query, Interface production
Detailed Financial Allocation Process
Executive Information Delivery
Extract and Transform
Extracting from legacy systems can often present a challenge to the most robust tools. Especially when source systems include complex data structures and high data volumes. Legacy systems often were designed to support monthly reporting cycles by accumulating and summarizing data before finally kicking out the hard copy report at the end of the month. Changing such systems to provide more recent information can make the difference between a successful information environment and simply a different format for the same old content. However, making sure the results are right can be a very expensive process.
PwC consultants used a new approach to attack this problem. For this client, they first created a program which understood these complex and unique legacy structures, which could be called by the Geneva ERS open API. This program opened up difficult legacy systems to the power of Geneva. Geneva was used to extract data from multiple sources. The Geneva development environment allows definition of source system data structures. Once defined to Geneva, business logic can be applied. The business logic is stored as a Geneva “View.” The Views are organized by logical frequencies like daily or weekly processes or on request processes. Once the environment was created, they focused on understanding the data structures instead of tracing through the entire legacy process. They used an iterative prototyping approach to discover the source of all the data contained in legacy downstream files. They successfully proved that the system could be converted from a monthly to a daily system. They extracted the source system data and populated the ODSs. The iterative prototyping approach used by PwC consultants shaved months off the delivery cycle and hundreds of man-hours spent in legacy system research.
The power of Geneva is evident in the types of source systems from which data is extracted. In addition to the complex legacy structure noted above, a DB/2 extract process reads 80 physical DB/2 partitions in parallel, scanning over 80 million records in approximately 10 minutes. All processing is completed in a fraction of the time it would take another data warehousing tool.
Load, Query and Interface Production
PwC set out to help create an environment that would allow many users, with diverse information needs to pull the desired information from the repositories. They created data and processing models that allowed all queries to be resolved in a single pass of the transaction detail. These models minimizes data storage requirements by eliminating duplicate data from transactions, but combining the data back together for report production in an efficient manner.
In the same pass of the data that loads the repositories, Geneva ERS also produces multiple reports and extract files for use by business users and other application systems. Many output formats are possible including reports, spreadsheets and files. And because the ODSs contain transaction level detail, users are able to choose what level of detail they wish to see. The data model also allows for use of Geneva’s date effective join capabilities. This enables users to create reports as of any point in time. Summaries can be created using organizational structures from any point in time.
The client choose to construct an ODS to support the ERP General Ledger being installed. However, the coding structure for the new ERP package differed significantly from the historical organization and account coding structure. The ODS supported all interface production, translating from old to new and from new to old. Global utilities were constructed that were called from the Geneva API. Because of Geneva’s ability to process the detailed transactions, all fields could be translated at the lowest level of detail. This enabled consistent answer set production for all interfaces.
The number of records loaded varies by repository, from 100,000 financial transactions (journal lines) to more than 1,000,000 policy transactions. In total, approximately 2 million transactions a day are added to the repositories. The single pass architecture even produces back ups using Geneva ERS in the same pass of the ODS. This daily execution of Geneva ERS reads over 450 million records in 23 different entities and 220 physical partitions. It writes over 900 million physical records to over 800 files. This process has run in 1 hour 3 minutes of wall clock time and 3 hours of CPU time.
Detailed Financial Allocation Process
The most sophisticated use of Geneva emulates an ERP financial cost allocation process. The insurance company recognized that with their volume of data, the ERP package could not handle the allocation process. They would have to change their business requirements or find another solution. They looked to PwC and Geneva to supply that solution. Client members and PwC consultants analyzed the allocation process and created custom programs which generate over 6,000 Geneva views based upon over 7,000 allocation rules. Geneva executes these views to scan the financial ODS selecting records eligible for allocation. It then allocates these costs through four allocation layers, such as products, and geographical units.
During the first year of implementation, the year-end allocation process read over 50 million records selecting nearly 3 million that were eligible for allocation. These 3 million records were exploded into 186 million allocation results. The process runs in 7½ hours wall clock time. The Geneva system produces these results 250 times faster than the ERP package. Because of this innovative Geneva solution, the business users were able to have their data represented exactly as they wanted.
Geneva ERS Executive Information System
Providing users with on-line access into repositories holding hundreds of millions of detailed records is a significant challenge. The PwC team developed an innovative approach to give users the access, but not compromise the performance of the ODS processes or require massive new processing capacity. The result was the Geneva ERS Executive Information System.
This system uses Geneva to produce summary “views” of the detailed transactions. These summaries were developed within Geneva, and could summarize by any field in the ODS. Approximately 20 different cuts of the data were developed. During the load and query processes, these queries are executed to refresh the summaries from the detail transactions. Because the summaries are always regenerated from the detail, no sophisticated update processes had to be developed and they also always contain the same consistent answer.
Users access the companies Intranet site, and select which summary to view. The Geneva ERS Java applet allows users to drill down within the report to lower and lower levels of detail. Because of the unique structure of the Geneva output file, data access if very efficient. This web site is accessible by over 300 business users
Most Geneva Views are created and maintained within the Geneva ViewBuilder software. This interface stores table structures, field types, relationships between structures, and the business logic to process the data. Geneva trainers trained over 100 company employees on site on the use of the ViewBuilder, for everything from basic view construction to use of the Geneva API, to key data warehousing concepts.
With PwC assistance the company implemented the first two ODSs. They have now moved on to developing additional warehouses on their own and multiple data marts. The result has been the ability to replace certain legacy systems with much more flexible architecture. The company has made major strides in meeting it objectives of “…a single, consistent source of financially balanced data to be used by multiple business users for decision support, data mining, and internal/external reporting.” They have been able to create business intelligence in an intelligent way.