FIG.06
PROJECT 03MLOps / Computer Vision
AI Image Caption Generator
A BLIP-based image captioning system enhanced by LLMs, with an Airflow-driven retraining pipeline.
ReactNode(Express.js)AWS ECS/ECRAirflowBLIP
Outcome:Automated retraining pipeline
Technical Focus
- Python
- Machine Learning
- Deep Learning
- Computer Vision
- Generative AI
- FastAPI
- React
- Next.js
- Node.js
- JavaScript
- MongoDB
- Firebase
- Docker
- AWS
Experience
[01]Professional roles and internships.
AI-ML Intern
2026FlyRankAI
- An ML pipeline for search ranking analysis - cleaning search data, building a baseline scoring rule, then training and evaluating models to prioritize which pages to review first.
Certifications
[05]Verified credentials and specializations. Click to view official verification.
DeepLearning.AI / CourseraNeural Networks and Deep Learning2024↗DeepLearning.AI / CourseraImproving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization2024↗DeepLearning.AI / CourseraStructuring Machine Learning Projects2024↗Packt / CourseraPython Programming for Quantum Computing2026↗Packt / CourseraMathematical Foundations and Quantum Mechanics Essentials2026↗
What I build at QAVREN
AI & Full Stack Engineering
Developing AI-powered applications and full stack products across machine learning, computer vision, generative AI, intelligent search, and modern web development.
Selected Work
[04]FIG.07
PROJECT 05Creator Economy / AI
Sponsorly
An intelligent sponsorship platform connecting creators and brands with automated matching, outreach pipelines, and campaign analytics.
ReactJavaScriptNode(Express.js)Tailwind CSS
Outcome:Automated brand sponsorship platform
FIG.06
PROJECT 07SEO / Search Intelligence
SEO ML Pipeline
An ML pipeline for search ranking analysis — cleaning search data, building a baseline scoring rule, then training and evaluating models to prioritize which pages to review first.
PythonScikit-learnDuckDB
Outcome:Ranked review queue with ~3x precision lift over baseline
FIG.02
PROJECT 08Mobile Development
PAK Homes
A native Android app for browsing and purchasing homes in Pakistan, built as part of a mobile development course assignment.
JavaAndroidFirebase
Outcome:Home listing & purchase flow for the Pakistani property market