GunduziAI is an AI-driven plant disease detection system that leverages the power of Deep Learning (Convolutional Neural Networks) designed to support farmers in identifying plant diseases (Tomatoes, Potatoes, Apple, Corn maize & Grape) and improving crop yield through timely diagnosis, expert consultations and treatment recommendations.

This Movie Recommendation System is a machine learning-based solution designed to offer personalized movie suggestions to users. It incorporates advanced techniques in data preprocessing, feature engineering, and ensemble learning algorithms to enhance the accuracy and effectiveness of movie recommendations.Using Python and machine learning algorithms (Random forest, SVM, Gradient Boosting) to create movie recommendations that basically suggests movies based on the available vast of pre-watched movie shows.

This project involved analyzing customer data to identify distinct segments based on purchasing behavior, demographics, and preferences. Using clustering algorithms (K-means, DBSCAN) and RFM analysis, we developed targeted marketing strategies that resulted in a 22% increase in customer engagement.

Technologies used: Python (Pandas, Scikit-learn), Tableau, SQL

An interactive Tableau dashboard that visualizes customer complaints data to identify trends, common issues, and resolution times.Through these visualizations, we can prioritize products with high complaint volumes, identify areas for process improvement, and monitor complaint trends to detect potential issues early on. By leveraging these insights, organizations can make data-driven decisions to enhance customer satisfaction and operational efficiency.Thus, the dashboard helped reduce complaint resolution time by 35% by highlighting bottlenecks in the process.

Features include:

• Complaint category breakdown
• Resolution time trends
• Geographical distribution
• Customer satisfaction scores

This project involved analyzing historical sales data to forecast future demand using various time series models including ARIMA, SARIMA, and Prophet. The final model achieved 92% accuracy in predicting monthly sales, enabling better inventory management.

Key aspects:

• Data cleaning and transformation
• Seasonality and trend analysis
• Model comparison and selection
• Performance evaluation (MAE, RMSE, MAPE)

This is a pest Identification Android App using Deep Learning Algorithms (Convolution Neural Network). Pestify is an Android application developed for pest identification using deep learning algorithms to help farmers quickly identify and treat pest infestations.

Key aspects:

• Data cleaning and transformation
• Feature Engineering
• Pest Model comparison and selection
• Performance evaluation (MAE, RMSE, MAPE)

This research examined the challenges and opportunities presented by the COVID-19 pandemic to online classes, focusing on the impact on educators and the educational system transformation.

Key Findings:

• Identified 5 major technological barriers to effective online teaching
• Discovered 3 unexpected pedagogical opportunities
• Developed framework for hybrid learning post-pandemic

This project involved web scraping Tesla's financial data from Yahoo Finance using R, followed by comprehensive time series analysis and visualization of stock performance metrics.

Technical Implementation:

Key Insights:

• Identified 3 distinct volatility patterns correlated with product announcements
• Developed predictive model with 87% accuracy for 30-day price movements
• Created automated reporting system for daily updates

A comprehensive analysis of FoodHub's customer order data to optimize operations and improve customer experience in the competitive food delivery market.

Business Impact:

An experimental design study examining the association between the Guided Care Model and unplanned hospitalizations among Illinois seniors with multiple chronic conditions (MCC).

Key Findings: