From deep learning models for medical screening to NLP pipelines for low-resource languages, I turn research into working, real-world software.
I'm an AI Engineer with a strong foundation in deep learning, computer vision, and natural language processing — passionate about building AI that solves real, human problems rather than staying confined to research notebooks.
What sets me apart isn't just technical skill — it's how I work. I'm known for being hardworking, disciplined, and deeply committed to whatever I take on. Once I'm responsible for a project or a team, I stay loyal to it, follow through fully, and don't cut corners — whether that's debugging a model at 2am or refining a report until it's genuinely ready.
Currently in my Final Year, pursuing a Bachelor's degree in Artificial Intelligence.
Currently working at O2Geeks, where I've contributed to NLP pipelines, prompt engineering, and REST API deployment — bridging model development with real production workflows.
Selected for an 18-month intensive English language and leadership development program sponsored by the U.S. Embassy's Regional English Language Office (RELO) — strengthening communication, critical thinking, and professional presentation skills.
Designing and building an intelligent mobile application for early detection and prediction of diabetic retinopathy using deep learning — the capstone of my AI degree.
An AI-powered mobile app that predicts diabetic retinopathy risk from retinal images using EfficientNet/MobileNet, with Grad-CAM explainability for transparent predictions, English/Urdu multilingual support, and downloadable medical reports for patients and clinicians.
CNN-based computer vision model detecting facial dark spots and blemishes with ~91% accuracy.
Multilingual spam classifier combining Naive Bayes, Logistic Regression, BiLSTM, and multilingual DistilBERT, with a Tkinter desktop GUI.
LSTM-based fake news detection system implemented from an IEEE research paper, with full preprocessing and evaluation pipeline.
Automated resume-to-job matching tool using TF-IDF vectorization and cosine similarity scoring.
Classical ML pipeline for classifying emails as spam or legitimate using text preprocessing and supervised learning.
Cross-platform task management app with AsyncStorage persistence, 5-screen navigation, and JSONPlaceholder API integration.
Route-optimization planner implementing A* and BFS search algorithms for efficient travel path-finding.
I'm actively looking for AI/ML opportunities — open to internships, collaborations, and full-time roles where I can apply deep learning and NLP to real problems.