What does it take to be a data scientist in the current time? Is it just about knowing a few tools and software? Saswata Kar, Director – Head of Data & Analytics GBS, Operations Leader, GBS Bangalore at Philips, shares his wisdom nuggets on what it means to be a data scientist, what scope this field offers, and where it currently falters in an interview with Analytics India Magazine.
Edited excerpts from the interview
AIM: How does your current role and responsibilities at Philips fit in with your analytics background?
Saswata Kar: I hold an Economics masters degree with a specialisation in Econometrics from JNU, post which I started working as a statistical modeller at my first job.
Currently, I am working with Philips, a leading player in the healthcare technology space. I am working with a team that helped in starting their finance analytics journey and establishing a single source of truth, streamlined reporting and analysis with near real-time availability of KPIs to drive right performance discussions and subsequently crafting the performance management. My contribution also includes providing guidance on determining which process and role types need to be centralised and building standardised systems.
AIM: What are some of the recent work you have been involved with?
Saswata Kar: My department is solving a wide range of operational problems in finance, supply chain, services and data management. They use descriptive, prescriptive and predictive models based on the needs of the hour. Some of them are really quick; some of them take a few weeks to develop. The most important point is to make a difference to the business – whether via increased visibility or simply better decision making, leading to improved savings.
AIM: What are the most popular tools and platforms used in the AI and analytics industry right now?
Saswata Kar: The AI industry’s commonly used platforms are SAP, SQL, Python, MS BI, Azure, QVDs & Python for Data Engineering, PowerBI (and other visual tools) for storytelling, Spyder & Jupyter Python IDEs for statistical modelling purposes. For a good business analyst, though, more than technology, good common sense is required. A good foundation of Lean methodologies, some training on project management tools like Scrum or Agile, and a good understanding of Six Sigma methodologies are helpful. Nowadays, organisations are moving away from point analytics or technology solutions to make overall processes more intelligent using AI. For implementing AI, various ranging tech stacks are required, right from RPA tools to intelligence document process tech as well as the above analytics tech stack.
AIM: What is the present state of data analytics?
Saswata Kar: It couldn’t have been better. Every individual, in some way or the other, is using data analytics to solve their wide-ranging real-life problems.
An entire ecosystem of various machines is reaching a certain maturity, which is making Data Analytics so sharp. State of the art yet inexpensive Internet of Things (IoT) ensures real-time collection of information, analysis, and recommendation to support people in making quick decisions and assisting them in performing tasks.
AIM: How do you think data science and analytics will evolve in the future?
Saswata Kar: One should take a step back and observe the overall IoT availability (where IoT stands for connected devices – electronics, software, sensors). One should start imagining what analytics can deliver on top of such an ocean of data, with a wide variety and incredible speed.
Considering the healthcare sector, in particular, on the usage of AI and analytics, I predict that hospitals in future will be equipped with connected technology – combining software, clinical decision support analytics and mobile connectivity – that will allow clinicians to monitor critically ill patients remotely, in real-time, to spot potential problems detected via data sciences early and act on them faster.
We are already moving towards a space where patients with chronic conditions could benefit from a set of connected monitoring devices that help them manage their condition in their daily environment, based on personalised feedback driven by data science pattern recognition. This is further supported by machine led coaching, which keeps the patients in close virtual contact with professional caregivers.
AIM: What are the main challenges facing the data science industry right now? What are the possible solutions?
Saswata Kar: One of the major challenges is the lack of diverse points of view in influencing outcomes of data sciences algorithms. We need people from all streams of society to contribute towards building decision sciences algorithms. With the advent of AI, this need becomes all the more important. Nowadays, organisations are looking at applications of technology in various parts of one process; eventually, different tech solves differing problems, and the full process becomes AI-driven. In such a case, there should be embedding diverse viewpoints within solutioning of each part of the process so that algorithms don’t throw only narrow points of view in which a process will operate.
It makes me nervous as to what will happen if the thinking person behind solutions is taken out, and we become over-reliant on technology. Data science can be a good enabler but should and can never be a solution itself. We definitely need to attract and appreciate various points of view, which is reflective of the diverse society we actually live in.
Data Science and Analytics are great enablers. If used rightly, it will solve many problems which actually cannot be solved by humans or are very expensive from a time commitment perspective—for example, having an experienced, fully attentive caregiver around a chronically ill patient’s side around the clock. But this can be made possible by IoT and Data Science’sScience’s combination, and caregivers can handle exceptions. They get a signal remotely via data-driven algorithms to attend to the patient in an emergency. The future is now!
AIM: What is your advice for aspirants who want to join this field?
Saswata Kar: Due to 24 /7 information dissemination, everyone knows that the data science industry is booming. Organisations are on a big-time hiring spree, as they understand what they need to build, most of the time. However, there is a misconception that the industry is just about knowing a few technologies, which can be easily acquired in reality. I have been acutely challenged while interviewing because everyone thinks data science is about knowing technology and algorithms and not about the problem that needs to be solved. Technology choice should depend on the problem type to be solved, not the other way. It was never about the algorithm or tech; it is always about applying common sense to solve someone’s real-life problem. I suggest everyone who aspires to be a good data scientist acquire a curious mindset (if they have not already) and ask five times why they are doing what they are doing or told to do. They shouldn’t be building a model just for the sake of the application of a particular algorithm.
(All the above views are personal and have no bearing whatsoever to the organisations that Saswata has worked with or working with.)
Subscribe to our NewsletterGet the latest updates and relevant offers by sharing your email.
I am a journalist with a postgraduate degree in computer network engineering. When not reading or writing, one can find me doodling away to my heart’s content.