Haroon Shoaib

Computer Science graduate from COMSATS University Islamabad with hands-on experience in machine learning, deep learning, data analysis, and web development. Built ML models on 8,000+ records achieving R2=0.9928. Trained neural networks using TensorFlow and Keras on standard benchmarks.

Proficient in Python, Pandas, NumPy, Scikit-learn, TensorFlow, Keras, SQL, Flask, and front-end web technologies. Seeking an opportunity to contribute to data-driven or technology teams and grow through real-world challenges.

Machine Learning Researcher at Independent Research Project (2026-03 – 2026-04)

Lifestyle Disease Prevention and Exercise Outcome Prediction

• Merged 3 public health datasets into a master dataset of 8,084 records, 20 features; trained and compared Linear Regression, Random Forest, and Gradient Boosting models.
• Random Forest achieved R2=0.9928, MAE=9.59 cal for calorie burn prediction and 83.3% accuracy (98% precision) for lifestyle disease risk classification.
• Key insight: workout frequency drives calorie burn (70.4% importance); BMI and weight drive disease risk (32.7% combined).

Android Developer, Final Year Project at COMSATS University Islamabad (2025-01 – 2025-12)

CareCircle - Android Parental Control Application

• Built a Kotlin/Android Studio app with real-time eye-blink detection and face-distance monitoring via camera-based logic; managed full development lifecycle independently.

Front-End Web Developer at Personal Project (2024-07 – 2024-08)

Gym Management System

• Designed and built a fully responsive website using HTML, CSS, JavaScript, and Bootstrap with membership management, class scheduling, and mobile-responsive UI.

B.S. in Computer Science – COMSATS University Islamabad, Attock (2022 – 2026)

Intermediate ICS – F.G. Degree College, PAC, Kamra Cantt (2020 – 2022)