Amazon Rufus (Search)- Rufus launch, Feb '24.- Developed personalization features on Rufus for customer tailored experiences- Developed Autocomplete, contributing 15% of total clicks on Rufus- Created a core data platform for A/B testing features across Rufus teams.

Alexa Shopping (Search)- Developed NLP algorithms to extract product insights from search and reviews, including classification, clustering, and recommender systems.- Implemented scalable data pipelines with AWS services like Glue, EMR, Step Functions, and SageMaker.- Led A/B testing for key features and translated findings into business insights for leadership.- As a Bar Raiser (Search Operational Excellence community), audited services across Amazon Search.

Amazon LastMile- Photo on Delivery (PoD)- Amazon Lockers- Developed Android + iOS App.- Created a UI Automation framework for the Android and iOS Apps to accelerate dev velocity.

- Conceptualized and designed a survey to collect data for research project based on personalized feedback for paper based Programming Exams- Analyzed collected data by calculating coherence, entropy, distinctiveness & coverage- Technologies used: HTML5/CSS, JavaScript, Weka, MTurk, Qualtrics, S3, Java

- Developed an end-to-end artificially intelligent grading system for an NSF funded project - Applied cluster analysis and Neural Network classification to develop the AI-grading algorithm- Computed results for DCC publication ‘Second Guessing – Designer Classification of Problem Definition Fragments’- Deployed the system for use in the Advanced Product Design (MAE 540) course, system currently in testing phase- Member of research team under Prof. Jami Shah- Tools & Methods: Word2Vec, Cluster Analysis, UMBC Similarity, Stanford Parser, WordNet, TF-IDF, Java