Aug 2023 – Jun 2025 · 1 yr 10 mos · Greater Chicago Area

Jan 2023 – Jul 2023 · 6 mos · Bengaluru, Karnataka, India · On-site

Jul 2020 – Jan 2023 · 2 yrs 6 mos · Bengaluru, Karnataka, India

• Fi is a FinTech startup founded by ex-Google Pay executives (Series C valuation ~650M USD, investors Sequoia, Ribbit Capital). As a core team member (seed-stage employee), responsible for scaling Data team, supervising data scientists. Hosted several townhalls for entire org; volunteered for back-to-office committee to manage transition from remote work for 200+ employees.
• 1) DemystiFi: Building the first-of-its kind AI financial assistant (key differentiator in Indian neo-banking space) for 2M+ users to enhance understanding of financial footprint, improve wealth management, and reduce friction on making payments, investing, applying for credit, analysing spends etc. Feature deployed on Day 0 of product launch in 2021.
• Driving cross functional collaboration between data science, product, engineering & design teams to strategise & build models which power the AI assistant. Directed improvement efforts for models that resulted in 4X increase in result relevance in just 6 months.
• Consulted with 10+ PMs across organisation to manage integration of AskFi with 20+ new product features within 1 year of launch.
• Overseeing development of new features for AskFi (autocomplete, spends analyser, related queries), increasing user engagement by 2X.
• 2) Spends Analyser: Built models to categorise user’s individual transactions for use in downstream products such as AskFi.
• As product owner, designed requirements, identified mechanisms to merge merchants across protocols & transaction methods. - Predicting a label for each transaction based on an internal category hierarchy using data points from real-time transaction info, historical payments, merchant information & properties, location-based information as well as user-provided labels & inputs.

Apr 2020 – Dec 2021 · 1 yr 8 mos · Bengaluru, Karnataka

• As an Industry Mentor, I supported and guided students that were part of Springboard’s Data Science Career track to transition into a career in Data Science. The role involved keeping track of each student's progress with the curriculum, discussing and evaluating assignments and projects, as well as guiding them to complete multiple capstone projects.

Nov 2017 – May 2020 · 2 yrs 6 mos · Bengaluru Area, India

• Gojek is SE Asia’s largest SuperApp providing 20+ services (like ride-sharing, food delivery, payments), founded by HBS graduate Nadeim Makarim (investors Sequoia, Tencent, Google). As first Data Science hire in India, I set up the function, mentored & scaled team to 10+. I conducted company-wide showcases on impact of AI projects across SE Asia, and was invited to speak to 1000+ folks at international conferences (Fifth Elephant, ODSC). Also founded & chaired Remote Work Committee that addressed remote work & collaboration across international geographies in 2018.
• 1) GoFood Search: Increased search-to-booking conversion for GoFood users by showing highly relevant search results & recommendations.
• Established AI-based approach to show relevant restaurants for a user’s search query, increasing search-to-booking conversions by 60%. - Worked closely with 20+ stakeholders in engineering, product & design teams to show personalised food recommendations & search results to 100M+ Gojek users, resulting in a 15% increase in search-to-booking conversions compared to non-personalised experience.
• 2) Pickup Point Prediction: Enabled pickup point selection feature by intelligently determining most popular pickup points at large public
• locations when booking ride on Gojek app; first of its kind in Indonesia (feature later launched by competitors like Uber, Grab).
• Reduced friction & for users & drivers, resulting in ~80% lower calls, chats between users & drivers, ~50% lower pickup time.
• Ran pilot for 1000 pickup locations in Jakarta. Seeing its success, deployed model across 1M+ locations across Indonesia & SE Asia.

Oct 2016 – Oct 2017 · 1 yr · Gurgaon, India

• Delhivery is India’s largest supply-chain & e-commerce logistics unicorn with 10K+ centres nationally (investors SoftBank, Tiger Global). I devised data-driven solutions to business and optimization problems in the logistics industry. Also led knowledge-sharing sessions on modelling techniques for logistics, and organised & hosted semiannual townhall for 400+ tech employees.
• 1) Reverse Logistics (RTO): Reduced reverse logistics cost for Delhivery’s largest client by identifying shipments having a high risk of return, prior to dispatch. Drove development of B2B analytics product, generating significant revenue for Delhivery in just first month of launch.
• 2) Vehicle Routing: Predicted volume of shipments at pick up client warehouses in top 5 Indian cities for maximum network utilisation,
• optimising routing of first-mile pickups through lower end-to-end delivery times & reduced operational costs from rerouting.

May 2014 – Jul 2014 · 2 mos · Pune Area, India

• Experimental High Energy Physics:
• Synopsis: This research project was computational in nature and involved the study of ‘jets’ (narrow beam of hadronized quarks) generated in high-energy proton collisions. Varied parameters such as pT Cut, Cone Size, Jet Algorithm, pTHat Cut, and analyzed and the effects the variations had on the various types of jets was studied.
• Technique Used: Monte Carlo Anlalysis using MadGraph, Pythia (event generators), Delphes (simulation and reconstruction software) and ROOT (data analysis software used at CERN); programmed using C++.

May 2013 – Aug 2013 · 3 mos · New Delhi Area, India

• While working in the Photonics Lab at the Department of Physics at IIT Delhi, I studied holographic techniques and recording material, and its potential application in solar panels. I varied the concentrations of acrylamide solution added during synthesis of the photopolymer solution, and analysed its effects with a varying exposure energy, while maximising efficiency and analysing viability of holographic-plate production methods.