A full-stack AI assistant that answers student questions using uploaded academic documents with retrieval-augmented generation.
GitHub Repository
DDoS Detection System (SDN + ML)
An ML-based intrusion detection workflow to identify suspicious traffic patterns and potential DDoS events.
What it does
- Processes network traffic features and classifies normal vs. malicious behavior.
- Generates interpretable output for model behavior and detection quality.
How it is built
- Data preprocessing and feature engineering with Pandas/NumPy.
- Training and evaluation using Scikit-learn classification models.
- Visualization with Matplotlib for distribution and detection insights.
Tech Stack: Python, Scikit-learn, Pandas, NumPy, Matplotlib
AI-Powered Document Verification System
A verification platform that analyzes uploaded identity/certificate documents and detects inconsistencies.
What it does
- Extracts text from document images using OCR.
- Compares extracted fields against user-provided data.
- Flags mismatches and possible tampering patterns.
How it is built
- Flask app for upload/verification routes and admin workflow.
- OpenCV preprocessing for better OCR quality.
- Tesseract OCR for text extraction from scanned images.
- MongoDB for document records, statuses, and audit-like tracking.
Tech Stack: Python, Flask, OpenCV, Tesseract OCR, MongoDB
Weather Mobile Application
A cross-platform Flutter app delivering real-time weather, hourly forecasts, and weekly forecast data.
What it does
- Fetches weather by current location and city search.
- Shows current conditions, 24-hour forecast, and multi-day trend.
- Handles loading/error states for smoother user experience.
How it is built
- Flutter UI with responsive layouts and Material components.
- REST API integration with async network calls and JSON parsing.
- State management for data flow and UI synchronization.
- Basic caching/offline strategy for better reliability.
Tech Stack: Flutter, Dart, REST API, Provider/BLoC concepts, Geolocation