Subarna Roy

Group Manager and Principal data scientist at IBM

I am a seasoned data scientist and AI/ML innovator at IBM having over 17+ years of experience in the field of data science, machine learning, and deep learning. As a senior technical leader, I have
worked with some of the biggest corporations and public sector solving their pressing business problems to come up with First-of-a-Kind (FOAK) solutions that eventually turned into successful AI/ML product. I have also published papers in international journals.

I hold a Ph.D. in applications of econometrics and game theory from IGIDR and was recognized as one of the top data scientists in India in 2018 by Analytics India Magazine.

All Sessions by Subarna Roy

  • Day 1

    April 22, 2021

  • 14:45  - 15:25
    Gartner says 80% of data science pilots will not deliver any business outcome and fail to go into production till 2022. Based on my 17+ years of experience with data science and AI projects, most failures happen when companies do not know what they are looking for from the data. It is of paramount importance that we take a structured approach towards building AI products right from the beginning. Early learners while understanding the nuts and bolts of data science, should also focus on acquiring knowledge and skills that help them build successful AI products that solve pressing business problems. In this talk I will focus on the Crisp -DM framework, elaborating on each of the components that go into building production-ready AI/ML taking examples from my past experiences and challenges. I have made few extensions to the Crisp-DM framework to highlight the importance of asking the right business questions at the beginning of a data science project. Also, I will provide few tips on how to engage with the clients, business users, and stakeholders at every phase of the project for the successful adoption of AI amongst them.
    Knowledge Sessions




Share

Subarna Roy