My journey to become an independent data scientist: Krishna Sai Vootla

Krishna Sai Vootla worked as a business analyst before becoming a data scientist. He graduated in Electrical Engineering from Indian Institute of Technology, Gandhinagar and worked for companies like, JPMorgan Chase and co., etc.

“When I joined my current company, the data warehouse was in a very bad state. Data from all sources was not integrated. Also, reporting was not automated. I made Lots of Data Engineering and ETL It was a rewarding experience implementing a complete, end-to-end warehouse (from API integrations, scripting automation, ETL, visualizations and deployments of ML code),” Krishna said.

Analytics India Magazine reached out to Krishna to understand his background in data science.


AIM: Is a degree in data science enough to get a job?

Krishna Sai Vootla: A diploma is not enough. I still believe in “slow and steady wins the race”. It helps if a person starts early. Institutions like IIT Hyderabad have a bachelor’s degree in AI. Although data science and ML have been around as a science for a long time, they have been put to good use over the past two to three decades. I believe that 90% of the research development has taken place in the last decade. Even if someone starts late, if they’re consistent, they can become an expert in a year or two.

OBJECTIVE: Tell us about your background in data science.

Krishna Sai Vootla: To be frank, I had no idea about data science until 2016. I joined a team working on a project for the Intel Embedded Systems Design Competition in 2016. We worked on a prototype to help teams emergency medical response to make quick and accurate diagnoses. I had no particular interest in programming or computing until now. But, I enjoyed developing pieces of code for my project. After tasting it, I wanted more, so I started to take on programming tasks. I started learning C++ in my spare time and was able to pick it up quickly. In just a few weeks, I learned the basics and contributed to the project. I taught myself programming, OOPS, Python, basic concepts of machine learning, SQL, etc. using MOOCs and academic blogs. Andrew Ng’s course is the best introduction to data science and machine learning. In his course, he covers many topics and how the transformation of technology can be helpful in various industries. I started to explore a lot of things and realized the tipping point we are in and understood the potential of data. In the past, decision making relied on a mix of domain knowledge and guesswork. Businesses today make informed, data-driven decisions. I’ve always been a numbers guy. Math and physics were my favorite subjects in high school. Practically, my job requires me to be on top of the numbers doing analytics to make data driven decisions for business growth.

AIM: What interesting projects did you work on during your studies?

Krishna Sai Vootla: Multimodal CNNs was a research project on using multiple input streams of images to see if it helped the model better understand. We found a dataset with forest scenes and trained a convolutional neural network to segment the scenes from these images. Additionally, we changed the architecture to add several visual modalities outside of RGB (like images from a depth camera). We found that adding multiple modalities of the same scene improved segmentation accuracy compared to only RGB images. Additional modalities proved to act as a new set of eyes for the model. It is difficult to understand whether neural networks, being a black box, have learned the correct criteria or not. For example, the model could associate anything with blue like sky if our training data has more images with blue sky and few images with blue lake. But with additional modalities like depth images, the model can differentiate these two; the value of the depth of a lake is always less while that of the sky is very high.

Telepresence robot: we used computer vision to decode human movements and worked to translate these movements directly to a robot.

Smart Medical Networks: We’ve been working on a prototype that could help emergency medical response teams make quick and accurate diagnoses.

AIM: How did you get your first job at Capgemini?

Krishna Sai Vootla: I got an internship on the Capgemini campus and joined an internal team as an ETL developer. My job forced me to learn the most important and employable skill I would use every day, SQL. I learned SQL in college, but working on an ETL team pushed me to learn and relearn all the details.

Then I moved to Tredence, a start-up at the time. While teaching myself ML and DL in college, I learned analytical skills like Excel (everyone underestimates this, but it’s really powerful for beginners to learn, although not scalable ), Table, Statistics, R Programming at Tredence. I worked on a customer segmentation project for Walmart. We have built a model to segment the 100 million customer base into different segments for the efficient allocation of marketing resources and maximization of cross-sell and up-sell opportunities. It gave me a lot of exposure and experience on how models are developed, deployed and how data science tools can help businesses grow.

AIM: Why did you choose to become independent?

Krishna Sai Vootla: Currently, I work as a Freelance Data Scientist at Toptal Talent Network, and my current client is Organifi LLC, a superfood company in San Diego, that I have been working with for over a year. My primary function is to gain insights and create KPIs that facilitate decision making based on high-level data. Recently, I built a sentiment and opinion extraction tool. I compared the accuracy results of different combinations of features and models. The tool ranked customer reviews and scored sentiment values ​​to identify weak spots in specific products and customer service. Fixing these issues led to steady growth in sales and fewer negative reviews over the following weeks.

My goal is to contribute to open-source / ML data science projects. I want to build tools, products for digital well-being. Also, I am working on some personal projects. Recently, I deployed a code on my PC that records my emotions every minute and stores them in a text file. After a month, I will summarize how my emotions change at any time of the day,” Krishna said.

AIM: How is the freelance scene for data science professionals?
Krishna Sai Vootla: There are plenty of opportunities outside of India. I think it is the right time for Indian software engineers, data scientists and IT professionals to go global like other countries. The Covid pandemic has definitely catalyzed this scenario. There are quite a few taboos in India about not having a stable and permanent job. I too faced some resistance and questioned/doubted myself many times during this trip. But my colleagues at Toptal gave me a lot of confidence. Now I’m at peace with the uncertainty of freelancing because I’m confident in my skills and I know there are a lot of companies out there that need a good data scientist.

Sean N. Ayres