razae User

I’m an AI undergraduate at COMSATS University Islamabad with hands-on experience in machine learning, NLP, and deploying production-ready AI applications. I focus on turning models into usable systems through REST APIs and practical automation.

In my AI & Machine Learning Intern role at DialerPortals, I supported data preprocessing, feature engineering, model training, evaluation, hyperparameter tuning, and performance optimization. I also collaborated in an Agile environment and contributed to documentation and workflow automation.

My projects include Titanic Survival Prediction (achieved ~78% test accuracy), Car Safety Prediction (85%+ accuracy with feature importance analysis), and Sentiment Analysis via Hugging Face transformer fine-tuning (improved F1-score using dataset balancing and learning rate scheduling). I’ve also deployed ML models with FastAPI (Swagger-documented endpoints) and built LLM-powered n8n automation workflows for extraction, classification, and report generation.