subject: Practical Approach To Build And Deploy Ppm [print this page] One of the key challenges faced by organizations striving to move from CMMI-DEV (or CMMI-SVC) ML2/ ML3 to CMMI-DEV (or CMMI-SVC) High Maturity (ML4 & ML5) is statistical process control. An important aspect of statistical process control is creating and using Process Prediction Models (PPMs). This paper presents a step by step approach to creating and rolling out Process Prediction Models.
Given below is a brief abstract of the paper is as given below:
"Building prediction models play key role in improving the organizational process maturity. Organizations face different challenges - Choice of outcomes, Scoping, Availability of data, Choice of leading indicators, data quality, Selection of right techniques to build the model, Validation & deployment of the model. This paper proposes a 5 step approach in rolling out prediction models at project as well as organization level based on Wipro's experience. This presentation also provides tips on Do's and Don'ts.
Building prediction models is key step in improving the decision making capabilities of an organization. Organizations face the below challenges in this journey:
Choice of outcomes
Scoping of the model
Availability of data
Choice of leading indicators
Selection of right techniques to build the model
Validation of model
Deployment of model
Every project will have multiple outcomes. There will be set of outcomes that are committed to the customer such as schedule, quality, resolution time and efforts. Other set of objective are internal focused such as profitability and utilization. It would be impossible to arrive at prediction models for each outcome. It is also important to upfront decide the applicability of the model in terms of projects, lifecycles, customers, technologies, unit of input as well as output measures. Once the outcome and input parameters are identified, a quick check needs to be done if the available data is homogeneous, reliable and sufficient to start the analysis."
Gurdarshan Singh Brar is working with Wipro as Head - Metrics & Data Management Group, Wipro and is a ME in Electrical and is a Master Black Belt, RBQNA Assessor, he has a lot of experience in Software Process Improvement.
Praveen Kurdukar is working in Wipro as Head Quality - Banking and Financial Services, holds a B E Electronics and Telecommunications degree and is a Six sigma Black Belt, Prince 2 Practitioner, ITIL Foundation, Managing Successful Programs (MSP Foundation), Certified Software Quality Analyst, Certified ISO 9001 Lead Auditor, Certified course leader for "Achieving superior service"
