Meta, Data Scientist – Facebook for Commerce/ Monetization –
New York City, NY May 2022 – Present
Reconsideration – Help engaged users reconsider purchases to drive advertiser value
Lead experimentation and measurement strategy for a 0🡪1 product. Helped successfully launch notifications for ads, multi ads and real time reconsideration units
Upleveled team’s measurement strategy and feature launches by conducting multiple types of AB tests
Identified long-term opportunities for ~$2Bn product investments and enhancements. Guided headcount decisions
Designed data-driven product strategy through multiple sizing analyses to prioritise engg resources, recommending goal metrics and team success criteria
Built feature engineering for a deep learning notification targeting model to target high intent users with shopping notifications
Amazon, Data Scientist 2 – Trans Initiatives
Seattle, WA
Aug 2020 – May 2022
Volume and Capacity Planning
Built an optimization model to determine feasible network allocations subject to capacities and constraints to improve coordination between different warehouses and connections.
Built and deployed machine learning Gradient boosted Tree (GBT) model to predict potential errors in planned volumes.
Developed a risk mitigation tool to evaluate feasibility of proposed warehouse and transport plans. This helped optimise transportation capacity during peak volume events
Designed and deployed a forecasting model to predict the ratio of processed packages to shipped packages
Visa inc, Data Scientist – Digital Products, Credit Modeling, Consumer Behavior – Predictive Products
San Francisco, CA Feb 2018 – Jan 2019
Global Travel Destination Prediction
Built a deep learning neural network using Keras to predict the likelihood of a cardholder travelling to a given city
Built a feature selection and importance algorithm to better understand how complex features affect predicted outcome
Deployed and productionalized using Scala Spark to score 1.2 Billion cards globally
E-commerce propensity
Developed a PCA based approach to identify merchant groups with similar e-commerce behaviours
Designed an RNN-based deep learning architecture to predict growth propensities for e-commerce behaviours
Built an opportunity sizing framework to estimate the incremental impact of the model on banks and merchants
Credit Risk – Small Businesses
Predict enterprise credit worthiness by leveraging visa transaction and customer data
Built regression-based modelling with extensive feature engineering and data efforts to link diverse sources and account for regulatory requirements
Implemented a normalisation algorithm to control for industry – geography effects for scalable deployment
Make titles and formatting better. It is not clear what is title and subtitle
Website to land on Experience as the first page