May 2019 – Jan 2020 · 8 mos · Noida, Uttar Pradesh, India · On-site

• Mentored ~50 students in Machine Learning, collaborating with GeeksforGeeks to design and deliver comprehensive classroom program, including the syllabus and labs.

Dec 2016 – May 2017 · 5 mos · Chennai, Tamil Nadu, India

Jul 2016 – Dec 2016 · 5 mos · Noida, UP

• Worked directly with the General Manager, Growth Analytics Team for the design and implementation of campaign management system.
• Worked in sync with the Betaout Team to carry out the integration of their marketing product with the current campaign management system.
• The projects undertaken as a part of growth analytics team include -
• 1. Campaign Manager, a machine learning model for customer segmentation and cross selling for various campaigns using python libraries including pandas, sklearn, scikit, matplotlib etc and classifiers like decision tree, naïve bayes, random forests, linear regression etc.The project also included a full-fledged monitoring system to monitor the performance of each campaign. The major campaigns covered under the same include -
• Failed Orders Campaign
• Due Date Campaign
• Recharge Date Prediction Campaign (using previous recharge logs,sms data)
• Dormant Users Campaign
• Cart Abandonment Campaign
• Cross Sell Campaigns etc.
• 2. BetaOut Integration, where I single-handedly carried out the integration of Campaign Manager with a third party tool BetaOut.
• 3. Promo Mapping, a generic multi-threaded model for mapping a customer segment to a particular promocode and for effective monitoring of the campaigns.

Jun 2015 – Jul 2016 · 1 yr 1 mo · Noida, UP

• Worked directly with the HOD, Risk Team for the design and implementation of fraud prevention engine.
• Awarded by the CEO, Vijay Shekhar Sharma with the ‘Legends Award’ in the Annual Town-Hall in Dec-2015, the annual award given to top 10% based on yearly performance, for excellent work in automation of fraud detection and prevention techniques.
• Awarded by the DGM, Fraud Prevention with the ‘Rising Star Award’ in Oct-2015, for the successful implementation of promotion abuse detector.
• Worked in sync with the Ali Pay Risk Team to strengthen our current Risk Engine, including Account Clustering Algorithm, Customer Rating Parameters, Email pattern based clustering etc.
• The projects undertaken as a part of the fraud team include -
• 1. Automated Re-seller detector for detection and prevention of promotion abuse in Paytm marketplace, that cancels fraud orders on a daily basis, with an accuracy of nearly 99%. (Python)
• 2. Advanced Address Normalization Algorithm, for mapping same addresses written differently. Its features include gibberish address token detector, address based spell corrector, etc. (R Language)
• 3. Integrated Negative List Management System, a graph based model to keep a track of the reason of an account getting blocked and performing a cascading rollback of a false positive. (Python)
• 4. Clustering Algorithm for clubbing together customer accounts belonging to the same individual / group of individuals, especially useful in systematically weeding out resellers, merchant buyer colluders, cc/dc fraudsters and nuisance customers to disallow them coming again on our system. (Spark Python)
• 5. Helping Risk Operations Team in Building Scripts and in generation of required Data for daily complaints.
• 6. Apart from these, analyzed various patterns in the behavior of the genuine customers and that of fraudsters on Paytm platform as viewed from various perspectives.