Key to success in data science: Domain expertise
A passion for data is a prerequisite to pursue a career in data science — datasets should instantly inspire you to infer, analyse and visualise information.
Organisations with large volumes of data crucial for their survival are constantly looking for talent that can dive deep into data piles to draw useful insights.
From humble statistics, the art of data interpretation has evolved into the complex multidisciplinary subject called data science, which calls for technological as well as human expertise to make meaning out of petabytes-sized data.
A passion for data is a prerequisite to pursue a career in data science — datasets should instantly inspire you to infer, analyse and visualise information. A successful career in data science also requires deep knowledge of mathematics, statistics, programming and AI techniques. Top that with domain
expertise — in, say, banking, insurance, healthcare or retail — and you will be raring to go.
Data science jobs involve handling large chunks of data with the help of computers. That calls for a sound base in programming. Data science students are generally taught the languages Python and R. “Today, the industry requires people having expertise in one language and knowledge of another language,” says Yogesh Kumar Bhatt, IT education and training vice-president at Manipal Prolearn, which offers a oneyear classroom diploma programme in data science.
Database management, data scraping and data wrangling are the three techniques covered by almost every comprehensive data science learning programme you find — both offline and online. Data scraping is the process wherein a technology sources information from a particular program, while wrangling is the process of cleaning and mapping raw data into a usable format.
Then there is exploratory data analysis, performed to spot patterns or anomalies, and data visualisation. Big data technologies like Hadoop and Hive, used to store and analyse big data, are also taught, besides an introduction to AI, with emphasis on TensorFlow, neural networks and image processing.
Companies may have interesting titles for their data professionals, but broadly classified, they work in two categories: One is the data scientist who has specialised in a domain, and the other, more of a data analyst, who is often paired with a domain expert.
Just as with other new sectors of knowledge, there is a shortage of people skilled in data science. Sridhar Gadhi, founder of digital solutions company Quantela, says the first challenge in hiring is the shortage of job-ready data scientists, owing to inadequate exposure to advanced skills. “The other challenge is the dearth of experienced data professionals in India. The ones with experience are mostly from traditional domains like finance and banking, and they find it difficult to adjust to our organisation, which is working on new technologies. We then look for candidates who have trained themselves in domains other than the traditional ones, through online courses or selfstudy,” he says.