Built a conversational hospital data assistant by turning local datasets into a Flask API, wrapping them with MCP tooling, and enabling natural-language querying through Claude for fast patient search, staff lookup, and weekly service insights.
A conversational task management assistant built using Python and LLMs. It lets you add, view, and delete tasks interactively through natural chat. Inspired by the ChatGPT Prompt Engineering for Developers course.
Built an interactive Power BI dashboard to analyze patient metrics, reducing reporting time by 40% and improving data-driven decision-making with 90% accuracy in patient trend predictions.
Developed Tableau dashboards with 95% accuracy in tracking sales trends and customer segmentation. Enhanced user experience with dynamic filtering, increasing data exploration efficiency by 50%.
Compared ML models in R (Linear Regression, Logistic Regression, Naive Bayes, SVM, and KNN Classifier), achieving 85% prediction accuracy and identifying the best model for heart attack risk estimation.
Preprocessed time series health data and used XGBoost to predict epidemic and recovery trends. Applied interpolation to handle missing values, resulting in a robust and accurate predictive model.
Implemented a sentiment analysis system using NLP techniques to classify user sentiment (positive, negative, or neutral) from social media, blogs, and reviews. Focused on real-time analysis and accurate text classification.