My Portfolio

Projects that demonstrate my expertise in Machine Learning, Deep Learning, Natural Language Processing, and Gen. AI. Each project is linked to its GitHub repository for further insights.

Smart Educator

Developed to proactively identify struggling students. This web app predicts exam results using ML, allowing early intervention to improve performance, and support even the most inefficient students.

Spam Classifier

Designed to classify messages. This app can be integrated with WhatsApp or SMS services, protecting users, especially in countries like Pakistan, from digital scams and unwanted spam, enhancing their security.

Verify User

Built for secure access, this face login system verifies users in real-time based on a pre-existing database, allowing or blocking access. It gets integrated with other systems to enhance security.

Car Price Predictor

Developed to estimate fair resale prices, this app helps buyers and sellers by predicting a car's value based on input parameters, ensuring informed and fair transactions in the used car market.

Duplicate Question Checker

Designed to identify similar questions, this deep learning-based tool detects duplicates, streamlining content management and improving user experience, especially in platforms like Quora.

Movies Recommender

Created for movie enthusiasts, this content-based recommender finds similar movies to the ones you love, making it easier to discover your next favorite film.

Audit Helper

Designed to ease the auditing process, this Python package automates the comparison of PDF documents with CSV data, reducing manual effort and error. It dynamically extracts and verifies information, stamping and saving PDFs for efficient, accurate audits.

Malik Land Ledger

Created for large landowners in Pakistan. This application maintains detailed land records, allowing Maliks to easily access crucial information about their properties and revenue. It offers essential details with potential for expanded functionality on demand.

Speak Ease

While learning the OpenAI API, I created this app for speech-to-text, text-to-speech, and audio processing, including Urdu. This experience enabled me to leverage the API effectively in other projects, guiding others in achieving their results with well-crafted prompts instead of fine-tuning.

Patient Records Extractor

Developed to streamline hospital records. This system uses computer vision (OpenCV) to automatically detect and extract ticked checkboxes from patient forms, saving time and reducing manual data entry errors.

Whatsapp Chat Analysis

This web app processes and analyzes group or one-on-one WhatsApp chats from a .txt file. It beautifully presents insights like member activity, message counts, busiest periods, and most common words, offering a deep understanding of chat dynamics.

source-code-not-available

Hand Written Digits Recognizer

Developed a handwritten digits recognizer using deep learning and the MNIST dataset. Uniquely, I tested the system with my own handwritten digits, achieving high accuracy with only one digit misclassification.

Diabeties Predictor

Built a diabetes prediction model using machine learning to analyze health data. This tool provides early risk detection, aiding in proactive management and lifestyle adjustments for better health outcomes.

Contact Patwari

An app to connect citizens with their local Patwari, providing accurate contact details from government sources. It simplifies access to the official, with future plans for automatic data updates and broader coverage.