Welcome to my portfolio, This page showcases a collection of hands-on projects that reflect my passion for turning raw data into actionable insights. From predictive models and NLP applications to interactive dashboards, AI-driven solutions and end-to-end data pipelines, each project demonstrates real-world problem-solving using tools like Python, SQL, GCP, Power BI, and machine learning frameworks.
This unique project demonstrates a customer care chatbot that blends AI text generation and pre-scripted decision trees to enhance digital customer service experiences. It reduced average response time by over 40% and improved user issue resolution accuracy by 30%, showcasing the power of hybrid automation in streamlining support workflows.
This amaizing project focuses on developing an automated loan default prediction model for LoanAnalytics Inc. using machine learning. It involves EDA, feature engineering, and model optimization to predict default risk, improve loan approval efficiency, and minimize financial losses by identifying high-risk borrowers early in the process. The final model achieved an accuracy of 92% and helped reduce potential default-related losses by up to 25%, streamlining the decision-making process for risk assessment and approvals.
This project analyzes IHS hospital component issues and wastage trends using Power BI. Key insights reveal strong correlations, seasonal patterns, and major wastage causes like Time Expiry (TIMEX). The findings provide data-driven recommendations to optimize inventory management and reduce wastage for IHSBSM. These insights contributed to a 15% reduction in expired inventory over six months and improved procurement planning accuracy by 20%.
This project uses historical data and advanced analytics to forecast real estate prices for Blue Ark Realty. By leveraging Tableau for visualization and uncovering key market drivers, it enabled data-driven decisions that improved pricing strategy accuracy by 25% and reduced property turnover time by 18%, helping buyers and sellers navigate the market more confidently.
This project analyzes sales performance across product categories at Ken’s Supplies using Tableau. By uncovering key trends and market opportunities, it delivered insights that improved inventory planning and boosted category-level profitability by 20%, enabling faster, data-driven decisions in a competitive market.
This project supports HELP International in identifying countries most in need by clustering socio-economic and health metrics using unsupervised machine learning. The analysis improved aid targeting accuracy by 30%, enabling more effective resource allocation to vulnerable populations.
In today's digital world, misinformation spreads faster than ever. This project presents a machine learning-powered Fake News Detection System that analyses news articles and classifies them as real or fake. By combining natural language processing (NLP) with ensemble learning techniques, It achieved 94% accuracy, helping media platforms and fact-checkers automate misinformation detection and reduce false content exposure.