This project builds an AI Dungeon Master for RPGs. It leverages Mistral/Llama2 to create dynamic narratives based on your choices (combat, character creation). Explore how the code guides the AI for a more immersive experience.
Predict customer churn (business metric) using ML. Analyzed usage, contracts & demographics to build a model (Python, Random Forest & Neural Networks) identifying at-risk customers for better retention strategies.
This project explores deep learning models to classify mango leaf diseases from images using a Kaggle dataset. Various models were investigated, including basic CNNs, EfficientNetB0, MobileNetV2, and finally Xception. Xception achieved the highest accuracy of 98.75% and was successfully saved for future use. Batch size optimization was crucial for Xception's training, as a larger size led to memory limitations.
This project involves a detailed sentiment analysis of British Airways customer reviews. By leveraging data extraction and predictive analytics, the project aims to unveil actionable insights on customer sentiment, which can be instrumental for enhancing service quality.
This project employs exploratory data analysis and K-Means Clustering on existing customer datasets to inform targeted marketing strategies for a Bicycle Company. By analyzing demographics, transactions, and customer behavior, we identify key customer segments that offer the highest revenue potential, guiding the company's marketing efforts for optimized engagement and profitability.
Developed an intricate Data Analytics Workflow Automation System, designed to streamline the collection, updating, processing, and reporting of business intelligence data. This system is encapsulated by a robust architecture that ensures efficient handling of large datasets, periodic updates, and user-friendly report generation.