As a student worker, I ensure that the technology in classrooms functions seamlessly. When issues arise, I provide in-person assistance or phone support to USC professors to resolve any technical problems.

As a Data Scientist at Persistent Systems, I conducted comprehensive data analysis on company products using predictive analytics for shipment date forecasting, delivering actionable insights to business stakeholders that helped drive growth and revenue strategies. I led data cleaning and exploratory data analysis (EDA) initiatives, uncovering key trends and presenting data-driven recommendations to clients, which enhanced satisfaction and secured a high-profit project. Additionally, I developed machine learning models for classification and named entity recognition (NER) for a healthcare company, enabling doctors to navigate medical protocol and drug testing/trial documents 60% faster. I also implemented Transformer models for natural language processing (NLP) to match movie synopses across multiple languages, including Spanish and German, reducing manual labeling time by 80%.

As a Machine Learning Intern at Persistent Systems, I led advanced research on Intrusion Detection Systems (IDS) and Phishing URL Detection, applying deep learning techniques such as stacked autoencoders, resulting in an 8% reduction in scam incidents. I published a high-impact paper titled "Comparative Analysis of Machine Learning Algorithms for Intrusion Detection Systems," contributing to advancements in the cybersecurity domain. Additionally, I designed and deployed anomaly detection models using OneClassSVM and stacked autoencoders, improving detection speed by 15% compared to human analysis and protecting critical digital assets from cyberattacks.

Worked on applications of Machine Learning in Cybersecurity

As a Natural Language Processing Intern at C-DAC, I researched Hindi language generation using LSTM, leading to deeper insights into linguistic structures and improving the accuracy of generated content by 20%. I enhanced the performance of BERT and LSTM models on Indian languages, addressing language-specific intricacies and achieving a 20% improvement in natural language understanding.

During my time as the Deputy Technical Head for TEDx, I led a team of three to develop a mobile application using Android Studio and Kotlin. The app was designed to enhance the event experience by facilitating real-time data storage and streaming for users and speakers through Google Cloud's Firebase platform. My role involved coordinating technical tasks, ensuring smooth collaboration within the team, and streamlining the development process to meet project deadlines. Through effective management and technical execution, we successfully delivered the app on time, providing a seamless experience for all event participants.