Introduction to Data Science Methodology for Business

  • Neena Sathi
  • Arvind Sathi

Your browser needs to be JavaScript capable to view this video

Try reloading this page, or reviewing your browser settings

You're watching a preview of subscription content. Log in to check access

Despite having great data scientists, most data science projects still fail due to a lack of proper skills and process to bring different teams together to manage and execute a data-driven data science engagement. This video will introduce 6 major steps of “enhanced” CRISP-DM methodology, which can be used in 1) selecting a data science use case, 2) preparing data sets, 3) developing data science models, and 4) setting up guard-rails in testing, training and deployment of models. It will appeal to students and professionals seeking to understand how to deal with large volumes of unstructured data.


This video provides an introduction to data science methodology for business and describes how organizations can automate their business processes.

About The Authors

Neena Sathi

Neena Sathi is a principal at Applied AI Institute. She has 30+ years of experience envisioning, designing, developing and implementing AI solutions associated with enhancing customer experience, back office automation and risk and compliance for many Fortune 100 organizations. She has worked as Director of AI Technologies at Carnegie Group, Inc, an AI startup, Accenture, KPMG, and IBM.

Neena has three masters degrees including MBA from leading US universities. She is Master certified integration architect from IBM and Open Group as well as certified Project management professional (PMP) from Project management institute. She is also certified in many Cloud and Cognitive technologies. She has widely presented and published many papers in AAAI, IEEE, WCF, ECF, IBM Information on Demand, IBM Insight, World of Watson, IBM Developer Works and various academic journals.

Arvind Sathi

Dr. Arvind Sathi is the Director of AI Literacy at KPMG and a faculty member with University of California where he teaches courses on AI and Analytics. Dr. Sathi received his Ph.D. in Artificial Intelligence from Carnegie Mellon University and worked under Nobel Prize-winner Dr. Herbert A. Simon. Dr. Sathi is a seasoned professional with leadership in Artificial Intelligence and Data Science solution development and delivery. Dr. Sathi was a pioneer in developing AI solutions at Carnegie Group (a Carnegie Mellon startup), leading to its successful public offering as a profitable AI company. At KPMG / Bearingpoint, he led the practices for Intelligent Process Automation, Enterprise Integration, & Analytics. At IBM, Dr. Sathi has led several AI and Data Science programs involving IBM products from IBM Watson, and Cloud business units, and has provided technical oversight to IBM’s strategic accounts. He has also delivered numerous workshops and presentations at industry conferences on technical subjects and holds four patents in information technology. He has published four books on analytics - Cognitive (Internet of) Things, Engaging Customers Using Big Data, Big Data Analytics, Customer Experience Analytics. He has also been a contributing author in a number of Data Governance books written by Sunil Soares and has published an article series on Advanced Analytics for IBM Developer Works.


About this video

Neena Sathi
Arvind Sathi
Online ISBN
Total duration
1 hr 26 min
Palgrave Macmillan
Copyright information
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021

Related content

Video Transcript


Introduction to Data Science Methodology for Business by Dr. Arvind Sathi and Neena Sathi. Data science grew through our experience with business intelligence or BI, a field that became popular in 1990s. However, the last 20 years have seen unprecedented improvement in our ability to take actions using artificial intelligence. This is still a very new area and there is no well-defined process or methodology, especially as it relates to guardrails, how do you manage and test and deploy your BI models.

Outline. CRISP-DM was the original methodology for doing the BI type projects. Today’s data science work deals with big data and many emerging AI technologies. As we adopt the BI technologies to AI deployments, these methodologies are facing many challenges. We will first describe these challenges, and then we will cover our modified methodology for data science work.

Our methodology extend our original methodology to include training, adaptive learning, and governance to the original methodology and adds the required AI related steps for knowledge engineering and management. We will also highlight the value of this methodology in the context of AI development and deployment.

After introduction of our enhanced seven step methodology in section two, we will describe each of these steps of our methodology in detail in section 3 to section 9. These steps include, one, describing use case. Step two, understand data. Step three, prepare data. Step four, develop model. Step five, evaluate model. Step six, deploy model. And finally, step seven, monitor model.

We will summarize video book content in the summary section 10. We will also summarize the key changes from original CRISP-DM in our enhanced methodology in summary section.

Audience. So who will benefit from this video book? This video book is designed primarily for executives managing data science or AI driven engagement, managers of business and IT organizations. This video book will also help anyone interested in exploring data science as a career.

And there are four possible candidates where this video can be used as introductory material. Number 1, data scientist who will use the video book to get initial exposure to CRISP-DM and AI modeling and deployment. Number two, data, AI, or automation engineers who will get enough understanding to use the methodology or embedded AI models in a larger system.

Test engineer will get an understanding of the specialized testing methods and processes needed for AI models. And finally, knowledge engineers can learn how the knowledge guideposts they have been developing gets packaged into an AI model and how it is then deployed.

Expected outcome. So what will you get out of this video book? You will be able to understand genesis of AI and analytics. You will be able to articulate data science process and methodology, understand the difference between BI versus AI. You will be able to articulate the major propositions of AI.

You will be able to understand how to analyze data sources and any enhancements needed. You will be able to understand how to build and integrate data and AI driven modeling techniques. You will be able to appreciate how AI can be used to benefit analytics in those cases. You will be able to explore how users and experts will be engaged for model measurements and monitoring. And finally, you will understand the controls and governance aspects.

Hi, I am Neena Sathi. I have over 30 years of AI development experience in many industries like telecom, health care, retail, and government. I have widely deployed AI applications in the US, Canada, and Latin America. I have widely presented and published AI related papers in many magazines and conferences. I have worked as a director of AI technologies at Accenture, CenturyLink, KPMG, and IBM.

Hi, I’m Arvind Sathi. I received my PhD from Carnegie Mellon University in the field of artificial intelligence. I’ve worked under Professor Herb Simon who received his Nobel Prize in economics and is also considered to be the father of artificial intelligence. I work today at University of California Irvine where I teach courses at a graduate level in AI and analytics and I also work at KPMG where I lead educational activities associated with AI literacy for 20,000 advisory consultants.

I have experience in developing and deploying AI systems in telecom, media, and accounting services. I have done this work across the globe in Australia, in Asia, in Europe, as well as in the US. I have written four books in the field of AI analytics and have several patents and many papers in the field.