Pawan Chowdhary

Model driven development for business performance management

Business process integration and monitoring provides an invaluable means for an enterprise to adapt to changing conditions. However, developing such applications using traditional methods is challenging because of the intrinsic complexity of integrating large-scale business processes and existing applications. Model Driven DevelopmentTM (MDDTM) is an approach to developing applications-from domain-specific models to platform-sensitive models-that bridges the gap between business processes and information technology. We describe the MDD framework and methodology used to create the IBM Business Performance Management (BPM) solution. We describe how we apply model-driven techniques to BPM and present a scenario from a pilot project in which these techniques were applied. Technical details on models and transformation are presented. Our framework uses and extends the IBM business observation metamodel and introduces a data warehouse metamodel and other platform-specific and transformational models. We discuss our lessons learned and present the general guidelines for using MDD to develop enterprise-scale applications.

A technical framework for sense-and-respond business management

In this paper we present a technical framework that supports sense and respond (SaR), the approach that enables an enterprise to adapt to a rapidly changing business environment. To implement the SaR approach, an enterprise proactively monitors trends and uses effective decision-support tools to help it act in a timely manner. We describe two pilot projects in which we implemented SaR prototypes and applied them to solve business problems. In the first pilot project we helped the IBM Microelectronics Division deploy an automated inventory management system based on our inventory optimization model. In the second pilot project, we helped the IBM Personal Computing Division deploy a SaR system in support of demand/supply conditioning. One of the components of this SaR system is an order trend model that provides early warning of constraints and excesses in the supply chain and helps make demand/supply conditioning more effective. Early results from these projects are encouraging and show that significant gains in profitability are possible.

Process information factory: a data management approach for enhancing business process intelligence

With access to critical performance indicators of business processes, executives, business managers and staff members can play a crucial role in improving the speed and effectiveness of an organizations business operations. The monitoring and analysis of business processes are complicated by the variety of organizational units and information systems involved in the execution of these processes. In this paper, we present a process information factory as a solution for managing performance data of business processes. The purpose of the process information factory is to provide a data foundation for a process-driven decision support system to monitor and improve business processes continuously.

Integrated model-driven dashboard development

Business performance modeling and model-driven business transformation are two research directions that are attracting much attention lately. In this study, we propose an approach for dashboard development that is model-driven and can be integrated with the business performance models. We adopt the business performance modeling framework, and we extend it in order to capture the reporting aspect of the business operation. We describe models that can effectively represent all the elements necessary for the business performance reporting process, and the interactions among them. We demonstrate how all these models can be combined and automatically generate the final solution. We further extend the proposed framework with mechanisms that can detect changes in the models and incrementally update the deployed solutions. Finally, we discuss our experience from the application of our technique in a real-world scenario. This case study shows that our technique can be efficiently applied to and handle changes in the underlying business models, delivering significant benefits in terms of both development time and flexibility.

Model-Driven Dashboards for Business Performance Reporting

Business performance modeling and model-driven business transformation are two research directions that are attracting much attention lately. In this study, we propose an approach for dashboard development that is model-driven and can be integrated with the business performance models. We adopt the business performance modeling framework, and we extend it in order to capture the reporting aspect of the business operation. We describe models that can effectively represent all the elements necessary for the business performance reporting process, and the interactions among them. We also demonstrate how all these models can be combined and automatically generate the final solution. Finally, we discuss our experience from the application of our technique in a real-world scenario. This case study shows that our technique can be efficiently applied to and handle changes in the underlying business models, delivering significant benefits in terms of both development time and flexibility

Modeling and simulation of building energy performance for portfolios of public buildings

In the U.S., commercial and residential buildings and their occupants consume more than 40% of total energy and are responsible for 45% of total greenhouse gas (GHG) emissions. Therefore, saving energy and costs, improving energy efficiency and reducing GHG emissions are key initiatives in many cities and municipalities and for building owners and operators. To reduce energy consumption in buildings, one needs to understand patterns of energy usage and heat transfer as well as characteristics of building structures, operations and occupant behaviors that influence energy consumption. We develop heat transfer inverse models and statistical models that describe how energy is consumed in commercial buildings, and simulate the impact of energy saving changes that can be made to commercial buildings including structural, operational, behavioral and weather changes, on energy consumption and GHG emissions. The analytic toolset identifies energy savings opportunities and quantifies the savings for a large portfolio of public buildings.

Enterprise integration and monitoring solution using active shared space

This paper describes on architecture and framework for business process transformation and monitoring, using active shared space along with business process solution composer, for achieving the above goals. This framework treats business data and business services as the first class citizens. The central notions of this framework are business artifacts and business services both of which users can exploit to define the key business data and performance indicators. There are producers and consumers of business artifacts and business services for the shared data space. Underlying services and data graphs can be configured in such a way that the service invocation and data mediation can be fully automated via the active shared space without human intervention. Active shared space advocates new programming paradigm that enables business level monitoring and business process execution based upon the definitions of business artifacts, business services and data graph. The architecture of the active shared space is detailed in this paper. A reference implementation is given for the sake of validation and discussion

Model Driven Data Warehousing for Business Performance Management

Traditional data warehouses are manually designed starting from specific requirements and anticipated data analysis needs. As a result there is frequently a disconnect between business process models, business definition of data artifacts and the data stored in the data warehouses as they are often designed manually and in isolation. Hence it has always been a challenge to keep the data warehouse in sync with the continuously changing business process models, resulting in both high maintenance costs and lost opportunities. In this paper, we present our model driven data warehousing (MDDW) approach in the area of business performance management (BPM). The purpose of MDDW is to bridge the gap between the business process models and the data warehouse models, thus enable the rapid adaptation to changes in the business environment. We describe our modeling framework comprising the various modeling elements and meta-models that capture both business and IT data artifacts

Business performance management system for CRM and sales execution

In 2004 the IBM Telesales organization launched a new customer segmentation process to improve profits, revenue growth and customer satisfaction. The challenges were to automatically monitor customer segment status to ensure results are in line with segment targets, and to automatically generate high-quality predictive analytical models to improve customer segmentation rules and management over time. This paper describes a software solution that combines business performance management with data mining techniques to provide a powerful combination of performance monitoring and proactive customer management in support of the new telesales business processes.

Automatic Discovery of Service Name Replacements Using Ledger Data

Recent studies have illustrated historical financial data could be used to predict future revenues and profits. Prediction models would be accurate when long-run data that traces back for multiple years is available. However, changes in service structures often result in alteration of the nomenclatures of the services, making the streams of financial transactions associated with affected services discontinue. Manually inquiring the history of changes can be tedious and unsuccessful especially in large companies. In this paper, we propose a machine learning based algorithm for automatically discovering service name replacements. The proposed methodology draws heterogeneous features from financial data available in most ledger databases, and hence is generalizable. Our proposed methodology is shown to be effective on ground-truth synthesized data generated from real-world IBM service delivery ledger database.