Software Developer — AI/ML Engineer
Like Hermione who mastered every spell in the book, I craft elegant code and intelligent systems — blending software engineering with the magic of Deep Learning and Computer Vision.
I'm Hafsa Salman, a passionate Software Engineer with hands-on experience building real-world solutions — from dynamic web applications and scalable backend systems to AI/ML-driven features.
I'm seeking a role where I can continuously learn, contribute meaningfully, and solve impactful problems. My expertise spans full-stack development, deep learning, computer vision, and NLP.
Currently building a Pakistan Sign Language AI Translator — making technology more accessible and human through the magic of computer vision.
Building intelligent systems with deep learning, MLOps workflows, and experiment tracking — from model training to production deployment.
Crafting robust APIs, scalable server architectures, and cloud-native systems across multiple languages and frameworks.
Illuminating user experience with beautiful, responsive interfaces — from React dashboards to interactive Streamlit ML apps.
Lifting projects into the cloud with CI/CD pipelines, containerization, and serverless sorcery on AWS infrastructure.
Specializing in Artificial Intelligence and Machine Learning. Building production-grade systems across computer vision, NLP, cloud infrastructure, and full-stack development.
Foundation in mathematics, physics, and computer science preparing for engineering studies at FAST NUCES.
An AI-based system translating between Pakistan Sign Language (PSL) and text. Features an encoder–decoder architecture for sign recognition, computer vision to extract hand gestures from video, and NLP to generate meaningful text output. Focused on real-world assistive applications.
Secure organ donor–receiver platform with role-based access for Admins, Donors, and Receivers. Automated matching via MySQL queries and serverless AWS Lambda, real-time email notifications via AWS SES, and RESTful APIs through API Gateway.
Role-based HMS for Admins, Doctors, and Receptionists. Implements patient data anonymization via hashing and encryption for GDPR compliance. MySQL triggers automatically log all CRUD operations ensuring full auditability and data integrity.
Real-time chat portal using Python and WebSockets with one-to-one messaging and FTP-based file sharing. Simulates and handles network issues like delay, loss, and corruption to test protocol reliability. Session-based authentication and user management.
A Harry Potter themed interactive portfolio built with Three.js, pure HTML/CSS/JS, and deployed on Cloudflare Pages. Magic is real.