Jul 2020 – Feb 2022 · 1 yr 7 mos

• Lead a team of Data Scientists to work on the following projects:
• Return to Origin (RTO) - Developed a classification model to predict the probability of RTO for a suborder. Build an API to estimate the probability in real time. The model is used for verifying every order at Snapdeal.
• Promised Date of Delivery (PDD) - Build a regression model to estimate the PDD for every shipment using historical lane features. The model is used for estimating PDD for every order at Snapdeal.
• Recommendation Engine - Used a combination of collaborative filtering and image similarity to generate similar recommendations for top products at Snapdeal. The recommendation engine is used to serve similar products on the product page.
• Product Deduplication - Developed a system to identify duplicate products using transfer learning and image similarity. Used it to remove duplicate items from homepage and search page which lead to a significant improvement in user experience.

May 2019 – Jul 2020 · 1 yr 2 mos

Dec 2017 – May 2019 · 1 yr 5 mos · Gurgaon, India

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

• 1. Was the sole owner of personalization and recommendations at ShopClues.
• 2. Personalized the homepage for ~10 Million users leading to a 100 % increase in CTR.
• 3. Created an in-house recommendation engine for up-sell and cross-sell using collaborative filtering.
• 4. The recommendation engine performed 10% better than the third party engine that was in place and contributes ~30% of the total orders for ShopClues.

Jul 2015 – Dec 2016 · 1 yr 5 mos · Gurgaon, India

• 1. Cleansing, validating, and manipulating complex and big data using R and Python.
• 2. Reshaping and merging multiple large datasets to obtain curtailed summary tables and listings.
• 3. Applying machine learning algorithms like Random Forest, Gradient Boosting, SVM, Nearest Neighbours etc. to solve complex data science problems.

Oct 2014 – Jun 2015 · 8 mos · Hyderabad

• To examine the effect of mobile telephony tower on village growth using data from different sources like Census, Economic Census, Household Amenities etc.
• Responsibilities:
• 1. To clean, reshape, and merge several data sets and make one master data set.
• 2. Apply difference in difference and propensity score matching on the processed data.

May 2014 – Jul 2014 · 2 mos · Bangalore

• Education and Employability in India
• Responsibilities:
• 1. Examine the key reasons that led to a decline in employability of engineering graduates in India.
• 2. Analyse the initiatives taken by various stakeholders to sustain the growth of Indian IT industry despite lack of skilled workforce.
• 3. Stakeholders include IT industry (providing technical, soft skill training), government, NASSCOM, and educational institutes.
• 4. Based on the above findings, developed a model to enhance the employability of an engineering graduate.