What’s all the Hoopla about Data Science, Data Analytics, Machine Learning, and Artificial Intelligence?

By Nitin Chandak

April 19, 2021

One day Ankita, pregnant for six months now, was browsing the Internet while sipping a cup of coffee. Among the litany of baby product advertisements, which she was getting used to, one particularly caught her eye. It was an application which claimed that it could tell how her baby would look like—right from the day they are born to their old age. She was amazed.

Such applications of AI are very common nowadays. In fact, your entire life rotates around AI, ML, and Analytics. You wake up and glance at your mobile: the facial recognition algorithm identifies you and unlocks your phone. You ask Alexa to turn on the light and play some music. Realizing who is speaking using voice recognition technology, Alexa plays from the personalized playlist. You talk to the remote of your SmartTV and it starts playing your favorite TV show right from where you have left it. It uses Analytics to segment you and on the basis of viewing history of others who belong to your segment, it even suggests some other shows you may like. Once done, you book a cab to go to office; you know the prices fluctuate due to demand. Google guides the driver of the cab to the destination. At the office, you mark your attendance by scanning your retina and log in to your system by scanning your fingerprint. I can go on and on and these are the very basics. We haven’t even come to the world of Social Media or the post-COVID transformations.

While some organizations have been exploiting this new-age technology from their very inception, a large number are now jumping on the bandwagon to transform the way they do business and derive competitive advantage from data. And both are hiring for such roles from business schools and engineering colleges in droves. It is, in fact, becoming so popular that a few institutes have started offering specialized courses in Analytics and Data Science.

What Do You Mean by Data Analytics, Data Science, Machine Learning, and Artificial Intelligence?

We all use Google, or Bing, for our search engine needs. On the basis of the earlier searches that you have done, your online activities, your demographics, etc., these search engines can create a profile of you and tag you to a customer segment.

Let’s say you are searching for a location you want to visit. As soon as you start typing the name of your destination, the engine would start prompting you various options. Once you have chosen your destination, the system will then show you the fastest route to that place.

Simply put, Data Analytics uses historical data on which analysis can be performed to derive actionable insights. It is used to answer questions that have been asked in a business context. In this case, the profile segmentation shows Analytics at work. There are four types of Data Analytics.

  • Descriptive—to understand what is happening or has happened in the past
  • Diagnostic—to identify why it is, or was, happening
  • Predictive—what can happen in the future
  • Prescriptive—what should be the future course of action

Artificial Intelligence is actually computers carrying out tasks which would otherwise require human judgement. In the instance cited earlier, it is about finding the fastest route between two points. It does not necessarily require any historical data or a predictive model; it just takes into consideration the distance between the two points and tells you the way without needing any input from you.

Machine Learning, a subset of Artificial Intelligence, does not require any human intervention but predicts the outcome of any event on the basis of the historical data and keeps learning as we feed more data into it, i.e., it is self-learning. In this case, Machine Learning (deep learning, a subset of Machine Learning which uses neural networks—a technique to replicate human brain) is behind the search engine predicting your destination. There are three types of learning problems in Machine Learning.

  • Supervised learning—where models are trained on labeled data
  • Unsupervised learning—where models are trained on unlabeled data
  • Reinforcement learning—where models learn from the feedback they receive

Data Science deals with the study of data for uncovering patterns, generating actionable business insights, predicting future events, amongst other things. While it may leverage AI/ML, there is more to Data Science than this, and all AI/ML is not Data Science. It is a broad and evolving field that is hard to define. However, the key to Data Science projects is the use of advanced algorithms which leverage Mathematics, Statistics, and an understanding of business context. Some common activities include framing the problem, data collection, data engineering, exploratory data analysis, model development, and insight generation.

                                                     Fig 1. AI vs ML vs Data Science

Excited? Next Question in Your Mind Is: “Would This Be a Good Career Option for Me?”

To answer this, let’s look at certain things that a career in Data Analytics/AI/ML demands:

  • Passion for Data: Having a passion for data, and turning data to information is a key requisite for having a successful career in this field. Data is proliferated everywhere (so much so that a term Big Data has been coined) in various forms—structured and unstructured. Data can be sourced in various formats—speech, text, image, etc.—and from various sources—calls, documents, emails, applications, databases, etc. In fact, majority of data are not even present in electronic formats. You should be able to use the available data to the best of your ability while ensuring that if there are opportunities to improve data collection, those should also be pursued.
  • Statistical Knowhow: Invariably, all the applications of Data Analytics require some understanding of statistics. Basic statistics is enough for most of the applications, but if you want to dive deep into technical details of Machine Learning algorithms, or want to develop some of these techniques, advanced knowledge would be required.
  • Knowledge of Machine Learning Algorithms: Having a knowhow of Machine Learning algorithms is a good thing. Whether a problem would need to apply basic techniques, such as Logistic Regression and Random Forest, or if it would require advanced techniques, such as Convolutional Neural Network, Recurrent Neural Network, Natural Language Processing and Computer Vision, a knowledge of Machine Learning algorithms helps speed up the process.
  • Grasp of Problem Statement and Business Requirement: It is of paramount importance that one understands the business context in which data is to be used. Same data can be used or represented in different ways for different contexts. Let’s say we have a collection of images of people involved in an accident; if the business problem is to identify faces in the images then we would use a different approach than the case where we have to identify the injury types.
  • Hands-on Experience in Coding: Business school graduates might not be expected to have hands-on experience in Data Science tools or software, such as Python, R, SQL, SAS, etc. However, it is always a good skill to have and goes a long way in ensuring that development-wise, your projects are on a sound ground.
  • Attention to Detail: It entails that we need to pay attention to where data comes from and what it represents. Paying attention to such details goes a long way in solving complex problems in an easier way, or eliminate them. When an organization wanted to automate one of their processes, they first started with extracting information from electronic PDFs to feed into another of their applications because two different teams were handling this. However, once they realized that the source of their information is already electronic, which they were converting into PDF format and sending from one team to another, they eliminated the entire process and populated data directly to the destination application.

What Are the Career Options in Data Science and Data Analytics?

As this is an evolving field, a lot of such opportunities keep on arising from different fields of business. However, there are some prominent areas where Data Science roles are offered to MBA and engineering graduates.

  • Pure Play Analytics/Data Science Organizations: Companies which focus only on Analytics or who have separate Analytics businesses hire from business schools for project management roles, in general, and sometimes as data scientists as well. They offer Analytics as a service to their clients. The project roles require one to:
    • act as a consultant, where you, remotely or on client site, study existing business processes and identify problems that may be solved using Data Science/Analytics projects,
    • act as a solutions architect, and ideate the approach toward solving problems and develop solution designs, and
    • act as a project manager to:
      • carry out cost-benefit analysis and create proposals and get internal and client stakeholder buy-ins,
      • manage seamless delivery of the project that sometimes requires hands-on coding, and
      • estimate the benefits derived from the projects.

Apart from this, one may be working on learning and implementing new technologies, going through regular up-skilling sessions, and exploring additional opportunities. While some projects might be technology-intensive, others may be reporting-driven. As one progresses in the organization, the focus shifts more from actual delivery to stakeholder management and business development.

The industries these companies may cater to are Insurance (Property and Casualty, Life and Annuities), Banking, Financial Services, Retail, Healthcare, Travel, Transport, Sports, etc.

  • Captive Analytics/Data Science roles: These are companies that hire for Analytics or Data Science roles for their in-house businesses from various sectors.
    • Banking and Financial Services Industry:
      • Quants Roles:
        • Trading Roles are offered to MBA graduates in high frequency trading firms where you will have to create algorithms which identify patterns in the secondary capital markets. This can then be used to take and exit positions.
        • Other roles involve creation of models for derivatives, structured financial products, and equity products which target risk management, new product development, business development, etc.
      • Credit Card and Banking Companies: These companies possess a lot of data about their customers including customer demographics, spending habits, credit scores, etc. The types of projects here would be on projects for business development, fraud analysis, customer segmentation, risk management, etc.
      • Insurance Companies: Insurance companies engage data scientists to help with regulatory compliance, fraud analytics, new product development, customer segmentation, underwriting support, risk management, leakage prevention, etc.

Skills required in all these cases are almost always hands-on experience with coding in Python/SAS/R/SQL, etc., along with sound understanding of financial markets, financial products, such as equities, derivatives, and the macroeconomic factors, among other things. While in some cases, MBA graduates are required to do hands-on work, in other cases their role is of project management.

  • Retail/E-commerce/E-gaming: The roles in the retail and e-commerce industries are in teams which deal with projects related to new product development, customer segmentation, cost optimization, customer experience improvement, market analysis, market segmentation, etc. A lot of such roles are as program managers where you would oversee the development and implementation of such projects/programs.

Apart from these, there are many fields, such as sports, real estate, telecom, ride-sharing, hospitality, etc., where data science and analytics are becoming key to their operations and corporate decision-making.

History is rife with examples of humans collecting and analyzing data to make decisions. With the advent of electronic format of data collection in the last few decades and the omnipresent Internet, this is surely just the beginning. I wonder what the future holds!


Nitin Chandak is a graduate from the Indian Institute of Management Calcutta, where he majored in Finance. He has been a consistent academic performer throughout. Post completion of his MBA, he has worked as an analytics consultant for three years. Prior to his MBA, he worked with Backspace (as Breakspace was then known) where he was instrumental in strengthening the Communications and Education Consulting divisions. Nitin is an ardent Manchester United fan and a chocoholic. He wishes to travel the world.

Our Pre-Placement Crasher program provides one-to-one/institute-level mentoring to students aspiring to join leading data analytics firms.  

Share on email