Four years back, in March 2018, something big was brewing in my life. Outside, it was a beautiful day – winters were finally going to rest, and spring was gradually taking its first breath. The roads were laden with new flowers, and people were basking in the sunshine that felt the right kind of warm.
On the inside, however, I could feel a storm underway. I was shifting from Mumbai to Gurugram – my first time interacting with the northern part of our country. I grew up in Vijaywada, Andhra Pradesh; went to college in Kharagpur, West Bengal; and did a small career stint in Mumbai, Maharashtra, for two years.
The move was one of those life-defining moments for me. I knew there would be a lot of change, but I was ready to give it my all. The idea of building the coolest startup in the country sounded way too exciting. After my interview with Arpit Dave, (co-founder of Runnr, a startup Zomato acquired in 2018), I knew I would be at the core of Tech that would redefine food delivery at Zomato.
I joined Zomato in the Analytics Team working closely with Logistics – the team responsible for getting food orders delivered to the customer’s doorstep through the vast network of 3,00,000+ delivery partners and the backbone of our food delivery business today. My job here was to improve the overall delivery partner experience.
In my first two weeks, I realised the problem statements I was looking to solve were not just new challenges, they were completely unique. Initially, I could not help but spend long hours working, and yet, I would not meet timelines. It was during this time I was asked to create a simulation product using python, which had to be replicated from the Go programming language. Since I came from an Economics background, it was difficult to understand anything that had code in it. This was a major challenge for me — codes seemed Latin to me. However, I was ready to learn Latin, if it came to that.
So I started – first, by constantly asking engineers around me the very basic questions. How do things work, which language is used, how can I learn them, etc? And then, I put those learnings into action. After a few months, I finally shipped my first simulation, which signals our platform to deactivate localities where the predicted demand of orders would exceed available delivery partners (to fulfil these orders), thereby leading to a poor customer experience. Eventually, this simulation became the basis for the final product that was launched! My joy knew no bounds. This experience encouraged me to learn more and solve new problems.
By the time I completed my first year, I had not only learnt the basics of all algorithms but was adept at the science behind Zomato’s Logistics product. I started searching for problems around the logistics framework that could best be solved using data science. This was when I stumbled upon building a model to estimate the credit limits for each delivery partner. I built the model end-to-end, and in the journey, learnt how models could make everyday life simpler. Once you know how a model works, all you have to do is put in the right features, and ta-da, you have a structure in place that can be scaled and replicated easily.
This is how my love and appetite for data science, models and big data grew. Subsequently, I realised I wanted to extend my core towards machine learning. So one day, I went to one of the tech leaders here at Zomato and asked if I could move to the Data Science Team. Thanks to the model I built recently and Zomato’s culture of letting people be driven by their passions, he instantly agreed.
While working in the Data Science Team, we had an inter-departmental meeting with the Risk Management Team to discuss how we can make food ordering and delivery safer for all our stakeholders – be it the customers, our delivery partners and the restaurant partners by identifying suspicious behaviour on the app. As we started working on this, we found a whole new rabbit hole. It was almost an awakening to the need to extend the scope further to stitch varied models to help reduce high-risk behaviours on the Zomato platform. And so, we expanded the scope of the Risk Management Team to many new areas – to use a wide range of app signals and behaviours to distinguish suspicious actors from genuine customers faster and more accurately.
Some of the projects that came out of this team have also helped improve the overall delivery partner and customer experience. One such project was to identify and club unique addresses together as a signal for our risk models. Today, the Logistics team also uses the same model to modify addresses for more accurate locations. Now, fewer delivery partners need to call customers to figure out the exact delivery location.
Overall, my journey in the Risk Management Team turned out to be everything and more I ever wanted it to be – a place I now call my second home. A great set of problems to work on and the best of colleagues (who became friends). I got to know some of the best people in Z – Rachna and Aanchal. A trio that came together for work, but stayed together for the shared love of food and hanging out.
I came to Zomato to discover the best of my professional capabilities but found way more. There’s something unique about working with people who are always up for going the extra mile and building products with the mindset that they will last forever. It’s here that I got to know how trusting people to do their best opens up so many growth avenues and that there will always be something more to dig into. In my four-year journey, I’ve realised that at Zomato, the need for an answer is greater than the need for mystery.
Even after spending 1,500+ days, it seems like the journey has only begun, and I can’t wait to find what’s in store.