Ask the Constitution
A full-stack retrieval-augmented generation (RAG) application that answers questions about the Constitution of Pakistan in plain language — and shows the exact articles every answer is grounded in.
Portfolio
Projects
AI and machine learning projects by Rayyan Khan — retrieval-augmented generation, computer vision, forecasting and applied LLM products. Every project below is real, most are live, and the featured ones have full case studies with approach, stack and honest limitations.
CottonDoctor-PK
A computer-vision tool that classifies cotton leaf diseases from a photograph, built to help Pakistan’s cotton farmers and agronomists spot crop disease earlier.
Lahore Smog Forecast
A machine-learning forecaster for Lahore’s air quality index (AQI), built to give residents and planners a data-driven short-term outlook during smog season.
Portfolio RAG Chatbot
A chatbot that answers questions about my work — projects, skills, experience — grounded in a curated knowledge base of my portfolio instead of a language model’s imagination.
RozgarMatch
A job-fit engine for fresh graduates: paste your skills, and RozgarMatch ranks live job listings by how well they actually fit — showing exactly which skills matched and which are missing.
PakBrief Agent
An AI agent that turns the day’s Pakistani news into a tight, readable briefing, powered by Groq’s gpt-oss-120b.
Conformer Speech Enhancement
A Conformer-architecture model that removes noise from speech recordings, trained and evaluated on noisy/clean speech pairs.
Fire & Weapon Detection
Real-time detection of fire and weapons in images and video with a YOLO-family model, reaching 0.89 mAP@0.5.
PDF AI Authorship Analyzer
Estimates whether the text inside a PDF was written by a human or generated by AI, from stylometric and model-based signals.
Roman Urdu Sentiment Analyzer
Sentiment classification for Roman Urdu — the way millions actually type — with a macro-F1 of 0.69 on held-out data.