Zarnab Rajpoot

AI Engineer with hands-on experience building AI systems end-to-end RAG pipelines, vector databases, and LLM-powered applications backed by strong full-stack skills in Python/FastAPI and React/Next.js/TypeScript. Recent work includes a realtime computer vision system for anomaly detection and a RAG-based support agent built with LangChain and Pinecone. Comfortable owning REST API design, authentication and permissions, structured JSON parsing, and document/image processing workflows.

ML Intern at FlyRank AI (2026-07 – Present)

• Built a production-grade RAG assistant for analyzing SEC 10-K filings, implementing document processing, embeddings, vector search, and LLM-based response generation over real financial data (e.g., Starbucks filings)
• Developed a real-time computer vision system for surveillance and anomaly detection using YOLOv8n for object detection and LTM-based logic for aggression detection, with a Python backend and React frontend
• Built AI/ML systems for diabetes prediction and real-time fraud detection, covering data preprocessing, feature engineering, model training, and risk scoring
• Designed and built a 3D interactive portfolio site using Three.js to showcase AI/ML work with responsive design and animations

AI/ML Intern at DecodeLabs (2026-06 – 2026-07)

• Built four ML/AI mini-projects in Python: a music recommendation system, a rule-based chatbot, a flower classification model using K-Nearest Neighbors (KNN), and a Cats vs. Dogs image classifier
• Applied supervised learning techniques (KNN) for multi-class classification and worked with image datasets for computer vision-based classification
• Handled data preparation across projects dataset cleaning, text labeling/tagging, and structuring data for training and evaluation

Bachelor of Software Engineering in Software Engineering – University of Lahore (2022 – 2026)

ICS in Physics – Punjab Group of Colleges (2020 – 2022)