Project Details

I designed and built a Power BI dashboard to bring together key business insights across sales, finance, marketing, and supply chain in one place. It analyzes over $3.74B in sales data and tracks important metrics like Net Sales, Gross Margin (38.08%), Net Profit (-13.98%), and Forecast Accuracy (81.17%).

To make the dashboard easy to explore, I added interactive filters for region, customer, segment, and time. It includes multiple views such as an executive summary, profit & loss breakdown, product and customer performance, and forecast accuracy tracking.

Through the analysis, I identified strong-performing regions like APAC and key revenue drivers such as notebooks, while also uncovering areas with profitability and operational challenges.

This project reflects my ability to work with data, build intuitive dashboards, and present insights in a clear way that helps stakeholders make better, data-driven decisions.

Live Interactive Dashboard

Project Details

Hospitality Domain Data Analysis is a data-driven project focused on extracting meaningful insights from hotel booking data. The project includes data cleaning, exploratory analysis, and visualization to understand trends in reservations, cancellations, and revenue. Using Python and data visualization libraries, key business patterns are identified to support better decision-making and performance analysis in the hospitality sector. The project demonstrates strong analytical thinking and practical data analysis skills through real-world datasets and visual insights.

Project Details

Developed an end-to-end Computer Vision solution for the automotive and insurance sectors to automate vehicle damage assessment. Using PyTorch and deep learning, the system processes exterior images to classify car conditions into six categories: Front/Rear Normal, Breakage, and Crushed.

The project covers the entire ML lifecycle, from image augmentation and preprocessing to model training and deployment. By leveraging transfer learning, the model achieves high-speed inference, providing utility for insurance claim automation and rental fleet inspections.

Technical Highlights:
Modeling: Built with PyTorch and optimized for robust multi-class classification.
Deployment: Interactive Streamlit application for real-time image uploads and predictions.
Business Value: Reduces manual labor, minimizes human error, and accelerates claim settlements through AI-driven insights.

Project Details

Credit Risk Modelling Using Classification is a machine learning project designed to predict the likelihood of loan default based on applicant financial and demographic information. The project involves data preprocessing, exploratory data analysis, feature engineering, and training classification models to identify high-risk and low-risk applicants. Various machine learning algorithms are applied and evaluated to select the most accurate model. A user-friendly web interface is developed using Streamlit, allowing users to enter customer details and receive real-time credit risk predictions. The system demonstrates how data-driven decision-making can support financial institutions in minimizing risk, improving loan approval processes, and enhancing overall credit assessment accuracy.

Project Details

Healthcare Premium Prediction Using Regression is a machine learning project designed to estimate medical insurance premiums based on personal and health-related factors. The project covers data cleaning, exploratory data analysis, feature engineering, and regression model development to understand the impact of different variables on insurance cost. Python libraries such as Pandas, NumPy, and Scikit-learn are used for data processing and model building. A user-friendly frontend has been developed using Streamlit, allowing users to input details and get real-time premium predictions. The application is deployed as a live demo for easy interaction and practical understanding of the model’s working.