Sobaan Nayyar

I’m an AI-focused Computer Science undergraduate at FAST-NUCES in Karachi, driven by building practical ML solutions that actually improve decision-making. I’ve developed a strong foundation through coursework in Data Structures, AI, Machine Learning, and Probability & Statistics, and I’m maintaining a CGPA of 3.57/4.00.

In my Data Science internship at Internee.pk, I built an NLP resume screening system using TF-IDF, cosine similarity, and spaCy to rank candidates by job-fit score and reduce manual shortlisting effort. I also created interactive ML dashboards in Flask and Streamlit for real-time analytics, and I validated engagement insights using an A/B testing simulation with a two-sample t-test.

I enjoy taking models from experimentation to deployment. Through projects like a wildfire risk digital twin, I engineered features from satellite layers, handled extreme class imbalance with SMOTE, tuned classifiers with GridSearchCV, and deployed an interactive risk map. I’ve also built retrieval-augmented agents that reason over job fit and explain match gaps using LangChain, Chroma, and a Next.js frontend.

Right now, I’m looking for a Data Science or ML Engineering role where I can ship end-to-end ML systems—pairing solid modeling with clean APIs, dashboards, and explainability. I’m especially interested in work involving NLP, agentic/multi-agent AI, and production-ready ML engineering.