Zahra.

Zahra Kazmi

Full-Stack Developer | AI Research Enthusiast | Aspiring Systems Architect

Overview

B.Sc. student at LUMS (2026) with a unique blend of full-stack development (MERN, FastAPI) and deep learning research. Experienced in building scalable financial tools at NETSOL and optimizing AI architectures at LUMS. Passionate about system optimization, mastering the product lifecycle, and thriving in mentorship-driven environments.

Lahore, Pakistan

Experience

May 2025 – Dec 2025

INTERNSHIP

Research Intern

Lahore University Of Management Sciences (LUMS)

Lahore, Pakistan

Conducted research on deep learning architectures for image segmentation, focusing on model optimization and performance evaluation using key accuracy metrics.

•Developed and trained deep learning models for image segmentation tasks, generating high-quality segmentation masks from dataset images.

•Implemented and compared multiple techniques and architectures to optimize model performance.

•Evaluated models using key metrics to measure accuracy and segmentation quality.

Key Achievements

→Optimized HitNet architecture for image segmentation on a novel, proprietary dataset.

→Generated high-quality segmentation masks using optimized deep learning architectures.

→Worked with a novel, proprietary dataset collected by the LUMS research team to train high-accuracy segmentation models.

→Generated high-quality segmentation masks using optimized deep learning architectures.

→Worked with a novel, proprietary dataset collected by the LUMS research team to train high-accuracy segmentation models.

PyTorchTensorFlowOpenCVScikit-learn

Jun 2025 – Jul 2025

WORK

Software Engineer Intern

NETSOL Technologies

Integrated internal financial APIs to build a reverse finance calculator, focusing on seamless user experience and scalable deployment.

•Developed a reverse finance calculator that resulted in a more efficient process for users to determine loan amounts and terms based on a target monthly payment, significantly simplifying complex financial decisions.

•Containerized and deployed the application, resulting in a stable and scalable environment that supported smoother development cycles.

Key Achievements

→Developed a reverse finance calculator that simplified complex financial decisions.

→Successfully containerized and deployed the application for improved scalability.

FastAPIReactPostgreSQLDockerAWS

Project Portfolio

PERSONAL

Present

Food Delivery Application

Developed a full-stack food delivery app using the MERN stack, implementing user authentication, cart functionality, order tracking, and dynamic restaurant menus to enhance user experience.

MongoDBExpressReactNode.jsMERN Stack

PERSONAL

Present

Multiplayer Connect-X Tic Tac Toe Game

Built a real-time two-player game with the MERN stack and Socket.IO. Implemented turn-based logic, room handling, and a custom win-detection algorithm for dynamic patterns (Connect-X).

MongoDBExpressReactNode.jsMERN StackSocket.IO