Shreyas Anand

Analysed past 5 year national cyberattacks data using Python libraries (Pandas, NumPy and Seaborn) and SQL to plot patterns indicating anomalies in network traffic, then implemented a LSTM model to predict cyber attacks with 95% accuracy
Developed data catalog and lineage documentation in collaboration with stakeholders to support data governance initiatives

Leveraged SQL to aggregate and clean over 2M healthcare records, ensuring data integrity for deep dive exploratory data analysis (EDA)
Analyzed 500K customer churn records using Python to identify key pain areas and lack of personalized recommendations
Pioneered interactive Tableau dashboards that translated complex user behavior metrics into visual narratives for executive stakeholders. Spearheaded data driven insights that empowered the team to increase customer retention by 20%, contributing to a $0.5M uplift in revenue

Constructed a robust deepfake detection framework utilizing Support Vector Machines (SVM) to differentiate between authentic and generative AI manipulated images
Refined image preprocessing and feature extraction pipelines, leading to a 10% improvement in classification performance compared to baseline models
Attained a detection accuracy exceeding 90% by fine tuning hyperparameters, ensuring high fidelity identification of synthetic media

Implemented deep learning models for news classification, personalized recommendation and summarization using NLP , LSTM, Bi-LSTM, TextRank and Sequential Attention achieving 93.4% accuracy
Presented the research at IEEE International Conference on Artificial Intelligence For Internet of Things