Subramanyam Reddy, Founder & CEO, KnowledgeHut UpGrad

Subramanyam Reddy (Subbu) is the Founder and CEO of KnowledgeHut, the leading technology workforce development company that helps enterprises and individuals around the world move forward to the next level with the help of technology. in 2012, at the age of 28, Subbu, as he is known to everyone, started KnowledgeHut. He on-boarded 12 employees and with an initial investment of $10,000 (Rs 5 Lakh). They almost gave up after the first two months of no revenues and no more savings to invest. But then the leads started flowing in and by the end of third month, they generated revenue of around $28,000 a(Rs14 Lakh). Since then, there has been no looking back. The company has grown from strength to strength and is now expected to clock an Annual Revenue Run-Rate of Rs300 crore in March 2022.


Data is a vast ocean of structured and unstructured information from a wide variety of sources. And a Data professional’s job is primarily to sift through it all and gather information and transform it into a form that can be easily understood, and then used by businesses. To achieve this, professionals like Data Scientists and Data Analysts must be well versed in data analysis, visualization, modelling, prototyping, and programming among other key areas. These essential functions lead to production and deployment of better products and services that solve real-world problems. 

A Data professional’s toolkit will consist of tools that enable them to explore, research, and analyze large volumes of data at high speeds and with accuracy, leading to impactful contributions to the project. Modern tools such as Matlab, SQL, Python, R and more are essential in the Data Science toolkit. 

Beyond these tools, data professionals need curiosity, the skill to gauge patterns, and the ability to collaborate with a cross section of functions to succeed at their job. 

Is Data Science a Career Worth Investing In?

It was just ten years ago that the Harvard Business Review quoted the then chief economist at Google saying that Data Science would be the sexiest job of the following decade. Today, a Data Scientist earns 25% more than a programmer, on average. 

The World Economic Forum in their Industry 4.0 insight report has tagged Data Scientist to the top job in 2022, which is no surprise considering how every industry depends on data-driven insights to develop products and services. The Cybercrime Magazine has projected that by 2025, the global data storage will reach 200 zettabytes. 

Competent Data professionals are the need of the hour to leverage this data and condense it into actionable information. There couldn’t be a better time to consider Data Science as a career. 

So, You Want to be a Data Scientist. Now What? 

A Data Scientist needs a significantly more varied skillset compared with other tech specialists. While tech skills do comprise of the essential skills that Data Scientists require, they’re not the only skills needed to be successful in the field. Other than familiarity with using programming languages for working with large datasets, you also need to be able to wrangle big data, and deploy your statistical, analytical, and business skills to use the insights from that data for aiding in decision-making that helps the enterprise grow. 

The key is to understand that the field of Data Science is as interesting as it is challenging, and it is also rewarding. Let’s look at the main skills you will need to master to thrive as a Data Scientist, whether you’re starting afresh in the field or switching to Data Science from another career path. 

Statistical Acumen

Statistics has been the precursor to modern-day Data Science with its penchant for finding patterns in data and continues to be relevant to how patterns in data are identified and then made available as usable information. 

Programming Skills

Programming is essential to Data Science primarily for th  e reason that an overwhelming majority of the data that Data Scientists will deal with comes from the digital realm. Termed Big Data for the sheer quantum of it, the data comes from various sources and there are powerful programming languages that make it possible to process it. 

Visualizing Data

Data, in simple terms, is information collected through usage across multiple touchpoints. This information by itself would make no sense to somebody who isn’t trained to understand it. When the quantum of such information increases, it becomes even more important to be able to visualize the important information from the data. Data visualization also makes it easier for people with no background in technology to be able to understand what the data is conveying. 

Understanding Data and Working with It

Like we discussed earlier, when we talk about data, we are referring to information. Consider buying clothes online. What you browse through is a data set, what you add to your wish list is another dataset, and what you finally end up buying is another. Each of these datasets contain several kinds of information ranging from the fabric to the brand to the size to the price and the shipping and the time of delivery promised. 

Let’s assume you abandon your cart without completing the purchase, that creates yet another dataset. The platform you are browsing apparel on would have thousands of customers visiting and creating such datasets. How would the platform make sense of these? What is popular with the customers? What kind of display strategy works better? What kind of price point, what shipping policy increases sales? 

Now consider other industries, healthcare, logistics, food & beverage, education, automobiles, finance, each of them has millions of such datasets that are collected in the form of numbers, values, and text that would confound anyone. This is the data that Data Scientists work with and here’s how they do that:

Data Collection and Storage

The raw data that is collected from various touchpoints is not suitable for analysis or visualization until it goes through several processes. The collection and storing of such data is the first step in working with it. 

Class Labelling

The splitting of the raw data into categories is the second step in working with data and creates more ease for further work.

Data Cleaning

This is one of the more tedious processes that Data Scientists encounter when working with raw data. This involves scanning data and getting rid of any anomalies such as missing values, incomplete fields, spelling errors, and anything else that makes the data unusable. 

Data Balancing

Considering that data is gathered across several different touch points, it is often the case that data is not balanced in terms of representative values. Data Scientists employ data balancing methods such as extracting observations from categories to balance the collected data.

Data Shuffling

This is one the most crucial steps in working with data. This is also the stage where patterns start to emerge. This is also the stage where statistical methods such as randomization and different sampling methods are applied to data to draw out patterns and predictions. 

Data Visualization

Finally, after several processes, the patterns drawn from the raw data can be presented in a visual form for everyone to understand and make decisions based on. 

Through all these processes, several tools and technologies play a crucial role helping Data Scientists wrangle, clean, slice and dice, and present data. Among them, R, Matlab, SQL, and Python are some of the most popular. Clean datasets make for more efficient algorithms. 

Looking to the Future

The US Bureau of Labour Statistics projects 11.5 million job opportunities in Data Science by the year 2026 with some of the highest salaries. For the last four years, Data Scientist has featured among the top three job roles in the USA on Glassdoor. 

The future sure looks bright for this essential function and people working to cultivate a strong skillset in the domain. 

Content Disclaimer

Related Articles