Murtaza Majid

Data Science graduate of Air University with 2 internships and 9 deployed ML and AI projects spanning computer vision, NLP, RAG pipelines, and tabular machine learning. Built and shipped production systems serving real users, from a citizen AI assistant (Awam Assist) to a 5-model marketing ML pipeline. Shipped a production MLOps pipeline on AWS EC2 with automated drift detection, GitHub Actions CI/CD, MLflow experiment tracking, and zero-touch Docker deployment.

Computer Vision & AI Intern - Scoutrex - Islamabad

(2025-12 - 2026-02)

• Optimized PyTorch models for cricket video analysis under motion blur and varied lighting; implemented MediaPipe pose extraction for biomechanical analysis and bowler speed estimation via FPS-normalized displacement.
• Engineered an action phase segmentation pipeline classifying bowling stages (run-up to follow-through) using temporal motion tracking and multi-angle camera synchronization.

LLM Task Trainer - Turing - Remote

(2025-11 - 2025-12)

• Engineered system/user prompts and multi-step task chains to optimize instruction following and integrate external API tools (Search, News, Yelp) using temporal logic.
• Mitigated hallucinations and improved model reasoning through rigorous RLHF evaluation, ensuring high-accuracy retrieval and context-aware response generation.

Data Science Intern - ITSOLERA Pvt. Ltd. - Islamabad

(2025-07 - 2025-09)

• Improved time-series forecast accuracy by 15% using LSTM, ARIMA, XGBoost, and Random Forest models on business metrics.
• Built interactive Streamlit dashboards to visualize trends, KPIs, and model predictions for non-technical stakeholders.

Bachelor of Science - Data Science - Air University (2026-06)