📍 Islamabad, Pakistan | 🎓 BS Computer Science, NUML (Grad. Jan 2027) | 🔭 Seeking fully-funded MS/PhD in AI/NLP
I'm an AI/ML Engineer focused on Retrieval-Augmented Generation, NLP, and applied machine learning. Currently an ML Engineering Intern at Flyrank AI, building ranking models on large-scale enterprise data. My academic and project work centers on making AI systems practical — from legal-tech RAG pipelines to deployed ML services.
- Research interests: RAG, Vector Search, LLM Systems, Predictive Modeling
- Currently building: E-Justice — a RAG-based legal assistant over 1,000+ legal statutes
- Actively applying to fully funded graduate programs (MEXT, DAAD, and others)
- Learning: Advanced NLP architectures, distributed ML systems
E-Justice — AI-Powered Legal Assistant (Academic Capstone) RAG system over 1,000+ legal sections (PPC, CrPC, Constitution) using Python, ChromaDB, and an LLM pipeline. Generates Urdu FIRs and bail petitions for non-English users.
Enterprise Content Decay Ranking Model Random Forest ranking model on 30,000 enterprise web pages (Flyrank AI), achieving 0.72 Precision@50 with grouped client holdout cross-validation — a ~2.5x precision lift over heuristic baselines.
Student Performance Predictor ML web app (Scikit-learn + FastAPI) predicting student outcomes from 7 demographic variables. Containerized with Docker and deployed on AWS ECS via ECR.
Electronics Shop Management System Cloud-based POS application (C# ASP.NET MVC, MySQL) with full CRUD across 7 operational modules, a Gemini-powered AI help desk, and a disaster recovery pipeline for catalog/customer data.
TradeMe/BidBud Data Extraction Suite Desktop automation tool (Python, PyQt5, Selenium) extracting structured product data across multi-page e-commerce sites, validated with 100+ integration tests.
Languages: Python · C# · Java · C++ · SQL AI/ML: Scikit-learn · NumPy · Pandas · RAG · ChromaDB · LLM APIs Backend & Cloud: FastAPI · Docker · AWS (ECS/ECR) · ASP.NET MVC Tools: Git · GitHub · MySQL · Selenium · BeautifulSoup