Smarter service delivery and business analytics      


Smarter service delivery and business analytics - Selected projects

Selected projects:

Application Assembly Optimization

Application Assembly Optimization is a delivery model that applies "factory floor" assembly and automation principles to the application development process. When coupled with a compelling collaborative environment and innovative measurements and incentives, it generates increased client value, revenue growth and cost savings throughout the application development life cycle. The AAO delivery model changes the face of global delivery, while raising the bar for superior value, higher quality, lower cost, shared risk, and productivity gains through a vested partnership with our clients. Research was a key partner with GBS in AAO since its inception in 2007. As IBM's patented and innovative delivery method of applying process automation to service delivery across the application development and maintenance lifecycle, AAO facilitates improved quality and productivity across 50 globally delivered services, resulting in faster and higher quality delivery of clients' applications. The AAO continues to refine and strengthen its global, common governance while leveraging local and global capabilities and skills from IBM strategic delivery locations across the globe. AAO's focus on outcome-based delivery, virtual collaboration, knowledge sharing and skills transfer, will also be a key differentiator for emerging and growth market organizations. In addition, AAO's ability to collaborate across delivery centers while diversifying primary delivery of key, high demand services from strategic locations across the globe, uniquely positions IBM as the optimal business partner to clients embarking upon complex, transformational engagements.


  • Scheduling Service Tickets in Shared Delivery
    Hari S. Gupta, Bikram Sengupta
    Intl. Conf. on Service-Oriented Computing (ICSOC), 2012
  • Who Do You Call? Problem Resolution Using Social Compute Units
    Bikram Sengupta, Anshu N Jain, Kamal Bhattacharya, Hong-Linh Truong, Schahram Dustdar
    Intl. Conf. on Service-Oriented Computing (ICSOC), 2012
  • SmartDispatch: Enabling Efficient Ticket Dispatch in an IT Service Environment
    Shivali Agarwal, Renuka Sindhgatta, Bikram Sengupta
    KDD, 2012
  • Work as a Service
    D. V. Oppenheim, L. R. Varshney, and Y.-M. Chee
    Ninth International Conference on Service Oriented Computing (ICSOC), 2011
  • Allegro: A Metrics Framework for Globally Distributed Service Delivery
    Daniel Oppenheim, Yi-Min Chee and Lav Varshney
    SRII Global Conference, 2012
  • Coordinating Distributed Operations
    Daniel Oppenheim, Saeed Bagheri, Krishna Ratakonda, and Yi-Min Chee
    Intl. Conf. on Service-Oriented Computing (ICSOC) 2011

Delivery Excellence Analytics

The focus of the Delivery Excellence Analytics project is two-fold; first on the development of consumable analytics that identify the potential performance issues and then help improve and enhance the delivery of IBM services, and second on the formulation of improved methods to estimate the resources and time required to deliver a project. Initial work the development of analytics has demonstrated success in identifying service contracts that may experience difficulties in the future from the characteristics of contracts. To date, several successful tools have been developed and deployed to predict future potential troubled projects based on the historical financial performance of the project, as well as attributes closely associated with a project's health. In the project estimation space, our efforts focus on developing improved methods to collect, verify, analyze, use historical data on a wide variety of IBM software project, and develop trend lines for different IOTs. These efforts have resulted in a reduction of losses through improved estimating and helps drive increased performance.

Analytics based on the available financial data examine multiple trend variables to gain insights into the future financial performance of service contracts, highlighting those which are suspected to experience a future decline. These approaches are able to detect subtle changes in the financial performance of a contract, giving risk managers’ and project managers’ valuable time to address problems. Attribute based methods use responses to a carefully designed set of questions developed using input from several experienced risk managers. Specifically, the questions focus on understanding project characteristics, such as the complexity and nature of the project, as well as assess the ability of the organization to deliver the project successfully. The purpose of these analytics is to identify patterns of trouble early into the engagement cycle, as well as gain a quantitative understanding of the scope of the potential problem during the development or delivery of services.

In project estimation, we are interested in not only improving estimation techniques, but also in practices that lead to better data collection, analysis of the data, and developing a deeper understanding of the nature of work and workers. Along the lines of improved estimation, we have developed tools to perform outlier removal, data re-normalization, integration of bottom up and top down estimation approaches, data collection, and generation of trend lines that relate required effort to project size. We have also investigated factors that impact global team productivity and performance, such as differences in sub-team time-zones, holidays, and languages. These factors can lead to reductions in team productivity, resulting in increased time, effort, and cost to complete a project.


  • On the Analysis of Global Team Performance
    Nianjun Zhou, Wesley M. Gifford, Krishna Ratakonda
    Services Research & Innovation Institute (SRII) Global Conference, San Jose, CA, July 2012
  • Predicting Transition Phase trouble improves Service Delivery Management
    Rose Williams, Welsey M. Gifford
    Frontiers in Service Conference, College Park, MD, June 2012
  • Risk Adjusted Pricing for Fixed Price IT Services Contracts
    Nianjun Zhou, Wesley M. Gifford, Krishna Ratakonda,
    Frontiers in Service Conference, College Park, MD, June 2012
  • Quantitative Modeling of Communication Cost for Global Service Delivery
    Nianjun Zhou, Qian Ma, Krishna Ratakonda
    IEEE Service Computing Conference (SCC), pp. 388-395, Sept. 2009
  • Variable Productivity Adjustment Estimation for Function Point Project Delivery
    Saeed Bagheri, Krishna Ratakonda, Rakesh Mohan
    IEEE International Conference on Data Mining Workshop, Sydney, Australia, Dec. 2010
  • An Effort Estimation Model in Project Delivery Using Hidden Setup Cost
    Saeed Bagheri, Nianjun Zhou, Krishna Ratakonda
    INFORMS Annual Meeting, Austin, TX, Nov. 2010
  • Accelerating Risk Management along Several Dimensions in Complex-IT Engagements
    Rose Williams, Krishna Ratakonda
    Frontiers in Service Conference, Oct. 2009

Analytics for Smarter Application Management Services

IBM Services Research has been developing a suite of advanced analytics tools that are designed to enhance IBM's Application Management Services (AMS) delivery process. The functionality of the suite spans a wide range of areas including ticket analysis, demand planning, resource planning, productivity management, and cost management. The accounts use the suite to perform five tasks: 1. measure and monitor key performance indicators, 2. detect trends, 3. identify available actions, 4. evaluate actions, 5. recommend change. Metrics that are monitored include average ticket volume and resolution times, resource utilization, service quality, etc. Once trends in key metrics are detected, various actions are identified, and analyzed by the suite. These actions include optimizing headcounts for application groups, identifying skill gaps and demand shortages, recommending best cross-skilling of resources, identifying areas for productivity improvement, providing and documenting knowledge to improve productivity, detecting and eliminating most common root causes, and automating resolution of most common ticket types. The tools have been in use by many AMS accounts. In order to speed up the deployment of the suite, IBM Research helped IBM GBS establish a new Center of Excellence within AMS practice and trained its members on how to use the tools. Since the establishment of the COE, the tools have been used in several dozen accounts. Currently, IBM Research is bringing the tools over a single platform on the web. Some of the functionality is already enabled on the web and the rest will be available soon.


  • IT Incident Management by Analyzing Incident Relations
    Rong (Emily) Liu, Juhnyoung Lee
    International Conference on Service Oriented Computing (ICSOC), Shanghai, China, Nov 2012
  • Improving Application Management Services through Optimal Clustering of Service Requests
    Ying Li, Kaan Katircioglu
    Frontiers in Service Conference, June 2012
  • Improving Application Management Services through Optimal Clustering of Service Requests
    Ying Li, Kaan Katircioglu
    SRII Global Conference, 2012
  • Statistical Modeling of Queuing Sytems for Capacity Management and Planning of IT Services
    Ta-Hsin Li and Kaan Katircioglu
    JSM 2012

IBM Strategic IP Insight Platform (SIIP)

SIIP applies deep analytics on over 30 million patent and scientific literatures from around the world to provide valuable insights into competitive landscape, white space, and IP portfolio. SIIP has unique capabilities to extract and analyze chemical and biological data (200+ million chemical compounds) to help Life Science & Pharmaceutical research and development.

Watch a video


Corporate Brand and Reputation Analysis (COBRA)

COBRA leverages advanced text and data analytics techniques to mine wide range of social media content, such as blogs, news, forums, and corporate internal information to derive customer and enterprise insights, such as brand and reputation insights, risk and compliance monitoring, market and competitive insights. COBRA is part of Cognos Content Insights (CCI) IBM commercial offering.


Equipment Health and Production Optimization Solutions

The goal of this project is to use advanced modeling and analytics to improve the performance and business impact by helping it drive lower equipment operational costs, improved equipment availability, longer component lifetimes, and optimized production levels.