Hi, my name is Muneera
I build AI systems that go from model to production.

End-to-end RAG pipeline that scores your CV against any job description in under 10 seconds. Built with a dual FAISS vector store architecture — one store for the CV, one for the JD — cross-referenced via similarity search and fed to Llama 3.3 70B via Groq. Returns a Pydantic-validated report with match score, skill gap analysis, tailored interview talking points, and rewritten CV bullets in the job's exact language. Zero external vector DB — fully local FAISS on CPU. Tags: Python · LangChain · FAISS · Groq · Llama 3.3 70B · Pydantic v2 · Streamlit · RAG · HuggingFace

Full-stack AI pipeline — paste any YouTube URL and get a structured summary in seconds. Built with a smart fallback architecture: youtube-transcript-api for instant caption extraction, Groq Whisper for videos with no captions, and LLaMA 3.3 70B for summarization. Supports multilingual videos, playlists, 4 summary styles, and in-memory caching. Tags: Python · FastAPI · React · Groq · LLaMA 3.3 · Whisper · Render · Vercel

An AI-assisted patient triage platform built with Node.js and Express, featuring an NLP-powered medical chatbot that provides context-aware health insights. Demonstrates full-stack delivery of an AI conversational layer in a production healthcare interface.