I am Kundan Pandey, a final-year Computer Science student and Data Science Intern focused on practical AI. I build machine-learning and analytics applications that take a problem from raw data through exploration, modeling, evaluation, and a clear user-facing result.
- Build classification and prediction workflows using Python, SQL, scikit-learn, and XGBoost
- Turn datasets into understandable dashboards and insights with Streamlit, Pandas, and visualization tools
- Create real-time interactive experiments with OpenCV, MediaPipe, and Pygame
A model is valuable only when its results are clear, reliable, and useful to people making decisions.
- End-to-end thinking - from data preparation and feature engineering to evaluation and presentation
- Practical delivery - ML projects built as dashboards, web applications, or interactive experiences
- Continuous learning - currently deepening ML deployment, model evaluation, and SQL analytics
| Area | What to expect |
|---|---|
| Live portfolio | A consolidated view of my technical profile, experience, projects, and downloadable resume. Visit portfolio |
| Machine learning apps | Prediction workflows with preprocessing, model evaluation, and user-friendly Streamlit interfaces. |
| Analytics projects | SQL analysis and dashboards that translate customer and operational data into business insights. |
| Computer vision | Gesture-driven interactive applications using OpenCV and MediaPipe. |
ML Deployment · Feature Engineering · Model Evaluation · SQL Analytics · Applied Computer Vision
| Project | Problem solved | Stack |
|---|---|---|
| Customer Churn Analytics | Churn prediction and analytics dashboard for surfacing customer-retention opportunities. | SQL, XGBoost, Streamlit |
| Credit Card Approval Prediction | Credit-approval prediction application built with a tuned logistic-regression pipeline. | Python, scikit-learn, Streamlit |
| Wine Quality Prediction | Web application for predicting wine quality with a random-forest model. | Python, scikit-learn |
| AI Fitness Coach | Personalized fitness and wellness platform with progress tracking and an AI coach. | JavaScript |
| Gesture Hill Climb | Webcam-controlled game powered by real-time hand-gesture recognition. | Python, OpenCV, MediaPipe, Pygame |
flowchart LR
A[Problem and Data] --> B[Explore and Prepare]
B --> C[Feature Engineering]
C --> D[Train and Evaluate]
D --> E[Dashboard or Application]
E --> F[Actionable Insight]
- Machine Learning with Python - IBM
- SQL and Relational Databases - IBM
- TCS iON Career Edge - Interview and Job Readiness Program
