I’m Omar Baig, a passionate Computer Science student at the University of Engineering and Technology, Lahore, currently maintaining a CGPA of 3.5. My academic and professional journey is centered around my deep interest in Machine Learning, Deep Learning, and Generative AI. I specialize in building intelligent systems that bridge the gap between complex AI models and real-world applications.
Through hands-on experience in both research and development, I’ve worked on impactful projects such as an AI-powered Medical Lab Report Interpreter and a Hybrid Movie Recommendation System. During my internship at Xavor Corporation as a Generative AI Engineer, I led the development of a mobility-assessment chatbot and automated Functional Reach Tests using MediaPipe and OpenCV—delivering medically validated solutions with over 90% accuracy.
I’m also skilled in NLP, Machine Learning, vector databases and deploying AI solutions using FastAPI and Flask. My goal is to contribute to cutting-edge AI systems that are interpretable, adaptive, and human-centric.
Passionate Computer Science student at the University of Engineering and Technology, Lahore, with a specialization in Machine Learning. Proven expertise in machine learning algorithms, and generative AI, backed by hands-on experience and multiple certifications that are also provided below in the certification section.
Xavor Corporation is a leading technology company based in Irvine, California.
Student of 5th semester
Below are my projects related to Machine Learning, Generative-AI, Recommendation System, DSA, Scrapping e.t.c
Developed an AI-powered Medical Lab Report Interpreter using OCR, LLaMA Vision, and vector search to translate complex medical data into simple, human-readable insights.
Developed a research framework for personalized LLM agents that combines preference graphs, multimodal feedback, and federated continual learning. Evaluated the system on synthetic datasets, demonstrating improved retrieval accuracy, personalization, and scalability compared to baseline methods.
Engineered a secure AI microservice for medical prediction using FastAPI, implementing model integrity via SHA-256 hashing and rate limiting to safeguard against tampering and DoS attacks.
Developed a hybrid movie recommendation system combining embedding-based similarity and deep learning with Neural Collaborative Filtering, improving personalization over traditional methods.
Built a modular Python pipeline that automates YouTube Shorts creation—covering script generation, TTS, stock footage integration, video editing, and publishing. The system leverages LLMs, APIs, and MoviePy to deliver scalable, reproducible, and fully automated short-form content.
Developed a powerful ML model with extensive data training. After meticulous feature scaling and data cleaning, I achieved an impressive accuracy of 83.37%—nearly matching the 1st place team's 83.4% in the Kaggle competition held in 2017.
Leveraged advanced data structures like graphs, trees, hash tables, and efficient sorting and searching algorithms (DSA) to create a standout car showroom management system.
Created a powerful Upwork Talent Scraper, integrating advanced sorting and searching algorithms for efficient data visualization and filtering.