• Developed an AI-powered quiz generation tool leveraging Google Gemini, LangChain, and ChromaDB, resulting in a 20% increase in user engagement and retention metrics.• Implemented a vector database using RAG to enhance the efficiency and accuracy of the quiz generation process, leading to a 30% reduction in processing time.• Revamped environment variable management system by transitioning from .env files to Google Cloud Secrets Manager, increasing security measures and reducing the risk of sensitive information exposure.• Automated feature deployment process by creating a CI/CD pipeline using GitHub actions and Docker, reducing feature staging time by 20%.

Utilizing Machine Learning for Enhanced Valuation of Personal Injury Claims | AI Innovation Project - Phoenix• Developing and optimizing ensemble machine learning model to accurately predict claim amounts for personal injury cases in the US, achieving a 70% accuracy rate. • Implementing XGBoost and LightGBM regression models for ensemble learning techniques to improve prediction accuracy by 20% compared to previous models. • Collaborated with legal domain experts to analyze key features and worked closely with cross-functional teams to enhance the preprocessing pipeline, resulting in a 10% reduction in overall processing time.Identifying Potential Areas for Urban Agriculture in Milan, Italy | Local Chapter Project• Implemented exploratory image analysis techniques like RGB Histogram analysis, Blurriness Detection, Total Energy, and Median Absolute Deviation using OpenCV and PIL to extract key features from Apple Leaf images, contributing to enhanced quality of the image dataset. • Analyzed tabular crop data using statistical techniques such as univariate, bivariate, and variance inflation factors to identify key features for optimal crop selection. • Utilized Python scripts to automate the webscraping process, extracting geospatial image data from Sentinel 2 via Google Earth Engine (GEE) and QGIS, resulting in the identification of 50 potential urban farming areas in Milan, Italy. • Led the coordinated collection of geospatial data from various online sources, resulting in a 25% increase in project efficiency and accuracy among a team of 100 global collaborators. • Developing customized interactive dashboards using Looker Studio to visualize urban farming areas, crop selection data, and smart pest management techniques, leading to a 20% increase in efficiency in decision-making processes.

• Implemented data processing techniques to preprocess and extract significant events from multi-channel EEG recordings, resulting in a 20% improvement in data accuracy.• Developed Spiking Neural Network (SNN) using snnTorch (PyTorch) for the classification of collected data, using encoded spike trains from the EEG data for training.• Utilized machine learning techniques to develop a novel feature extraction method for significant event detection in EEG signals, achieving an overall accuracy rate of 55% and reducing computational resources by 25%.

Department of CommunicationCMN 174 - Social Media Communications (Spring 2023)CMN 131 - Introduction to Public Relations (Fall 2022, Winter 2023)CMN 001 - Introduction to Public Speaking (Spring 2022)Department of Mathematics and StatisticsMAT 021D - Vector Analysis (Winter 2022)MAT 021A - Calculus (Fall 2021)

• Developed and executed a comprehensive data mining plan, resulting in the creation of datasets for research in developing deep learning-based classifiers for Audio Event Detection (AED) and adversarial attacks on AED classifiers.• Implemented advanced techniques for audio data augmentation using Python and Audacity API, resulting in an increase of accuracy by 10% for the deep learning-based classifiers.• Contributed towards two published papers showcasing disruptive and adversarial attacks on Audio Event Detection (AED) deep neural networks, demonstrating expertise in experimental design and machine learning techniques.

•Part of the student team that helps incoming freshmen adjust to college life and achieve success in both academic and social aspects of their life.•Collaborated with more than 50 Peer Academic Leaders to help improve the program for freshmen.