- Being a data scientist requires one to be a “jack of all trades” and be able to work from scratch to work out strategic insights.
- Four key skills required in data scientists: programming / hacking; mathematics and statistics; business acumen; creative storytelling
- Businesses want data scientists to communicate their findings or solutions through engaging and relevant stories.
Data science — the job Harvard Business Review called the sexiest job of the 21st century — continues to play a massive role in the transformation of various industries. Amidst the uncertainty of job losses across IT and other industries in India, data science is one skill which is expected to provide only increased job prospects with each passing day.
However, there is often a deep disconnect between the aspirations of data scientists and reality. Often, data science is less about complex modelling and machine learning and more about data discovery and “wrangling”. On the other hand, companies wanting to use data science to surface impactful insights often discover aspirants whose mastery doesn’t extend beyond some popular tools.
FactorDaily’s Date with Data Science event (part of our #FutureOfJobs initiative) was meant to kickstart a realistic conversation between experts and data science aspirants and enthusiasts about how to get started in this field. The event was organised at the FactorDaily office in Indiranagar on March 18.
Guest speakers Achint Thomas (principal data scientist at Embibe), Bharath Cheluvaraju (chief computer scientist at SigTuple), Puneet Mathur (analytics and decision science leader at Target India) and Guruprakash Sivabalan (co-founder of Xobin) put forth and discussed their views on data science, the key skills that are important for a data scientist and how to get started — both as an aspiring data scientist and as a recruiter.
Here are the key takeaways from the event:
- Our generic view of data science is not productive. While fundamental skills are relevant across industries, real contributions can be made only by employing data science in the context of each business or industry.
- Being a data scientist requires one to be a “jack of all trades” and be able to work from ground up all the way to surfacing strategic insights.
- Inquisitiveness or constantly seeking answers to questions using data is the fundamental stepping stone to being a good data scientist.
- Four key skills were repeatedly highlighted by data scientists from the industry: programming / hacking; mathematics and statistics; business acumen; creative storytelling.
- In India, we have a large pool of talent with programming / hacking skills, but with only moderate math skills. Experts feel the need for more talent with deeper understanding on mathematics.
- More importantly, the biggest gap is in business knowledge to employ data science to solve problems in different industries. While this knowledge comes with experience, the inquisitiveness to ask the right questions and look for answers is seen as an important prerequisite to develop this skill.
- This is also reflected in how well data scientists understand the tools and modules they are using. For instance, using a Python library to manipulate data is not data science. Real solutions can emerge only when there is a deeper understanding of how those libraries work and whether they can be tweaked to solve new problems.
- Data scientists have to be storytellers. Gone are the days when you can work in a silo with your headphones on. Businesses want data scientists to communicate their findings or solutions through engaging and relevant stories in order to be able to use them to solve problems.
- For an aspiring data scientist, finding a mentor is important, although it is easier said than done. Getting hands on with data is the fastest way to get into the field.
- Similarly, it’s important to engage with the larger data science community through forums (like Kaggle) and work on interesting problems.
Subscribe to FactorDaily
Our daily brief keeps thousands of readers ahead of the curve. More signals, less noise.