As a Machine Learning researcher, Neelanshi Varia thinks through alternatives for mathematics, utilization of available data, and fine-tuning models and applications.

Can you share the most interesting part of your career journey?

I began my industry career at Deloitte, and I can confidently say that it is the best place I could have started my career at for multiple reasons. First, being able to translate my research and technical skills into state-of-the-art industry applications is such a rewarding and thrilling process. Second, doing it while simultaneously learning new technologies and brainstorming innovative applications is great. And seeing the impact that these projects have, my motivation is constantly reinforced to solve more real-life problems through AI. Additionally, getting exposure to different industries and staying on the top of the ever-evolving landscape of AI in different domains is very useful.

And seeing the impact that these projects have, my motivation is constantly reinforced to solve more real-life problems through AI.

Thought leadership and various knowledge repositories give me a better perspective of the current state and where AI is going to be in the next few years. I have been a Machine Learning researcher for almost 4.5 years now and thinking through ways of improvisation in terms of mathematics, better utilization of available data, fine-tuning models and areas of applications taught me a lot of skills. Being able to apply those ideas from papers on actual client problems and to Deloitte’s AI products is amazing.

What excites you most about working with data and AI?

AI is at its peak ever since the AI winter and as some say, next industrial revolution, thanks to advanced compute and storage resources. I feel working with data is like working at a gold mine and being equipped with AI—it’s always like finding treasure. I like to think about how different data points that might seem unrelatable to a human eye can be exercised into making an existing or new system faster and better.

It is exciting to see the applications augmenting capabilities of humans in practically all the areas possible.

AI is omnipresent, and to be able to create interesting applications just with the right data and tools is great. I have been able to apply data science on projects ranging from micromanaging growth of a home plant to deciding agriculture strategies using satellite data, from managing tasks of associates in a store to managing supply chain of products across a whole country, from detecting objects on the road while driving a car to detecting roadmaps, traffic, and land types from aerial view. I love the versatility of data and AI.

Describe an interesting project that you have worked on.

I worked on creating a tool that uses data across multiple universes to identify a patient and market appropriate drugs or action items based on their journey. Using behavioral, marketing, health, claims, and other data points, the goal was to identify if a person could possibly have a certain disease and, based on where they are in their disease identification journey, recommend appropriate drugs.

It was thrilling to work with such a ubiquitous data set, to experiment with it and to make it into something that will have a significant impact in the industry.

We created a patient-wide view across all parameters and the team developed several AI frameworks for different use cases. This tool is one of the first of its kind. Something similar is available for health practitioners in the market, but not for patients. It was thrilling to work with such a ubiquitous data set, to experiment with it and to make it into something that will have a significant impact in the industry.