•Developed, monitored, and enhanced the performance of a popularity-based machine learning model for predicting trends.•Constructed a recommendation system using collaborative filtering to strengthen user personalization.•Proficient in Python, utilized for building microservices that orchestrate result assembly with layered personalization.•Architected numerous widgets and microservices, optimizing cost and time efficiencies across application layers and services like Redis.•Implemented pipelines in Azure Databricks using PySpark to manage extensive data volumes, including threading, database querying, and integration with platforms like Cosmos DB, data warehousing, and Solr.•Engineered tailored applications for predictive analytics, automated decision-making, and optimization algorithms.

Conducted a study on minimalist attacks on RL agents, demonstrating that just 1% of state attacks could reduce performance by 40%. Additionally, developed and validated a resource-efficient defense against patch-based attacks on monocular depth estimators, achieving a 30% reduction in RMSE error. This was implemented using Python with TensorFlow and PyTorch frameworks.

Completed an orientation covering Python and Machine Learning fundamentals, including data preprocessing, data visualization, and ML algorithms. Developed a two-stage Flight Delay Prediction System utilizing Machine Learning techniques.

Developed a mobile application designed to enhance children’s English language learning through music. The app challenges kids to select the next word in songs from multiple options that appear on the screen, promoting hand-eye coordination, listening skills, and fluency development.