Building AI-Powered Applications That Solve Real Business Problems
Computer Science student specializing in AI-powered web applications, automation systems, computer vision workflows, full-stack engineering, and modern AI-assisted development.
3+
AI Products
FSD
Pakistan
UAF
CS Student
Open for AI Engineering Roles
Faisalabad & Remote Available
AI/ML-focused Computer Science student building production-ready web platforms.
I am a Computer Science student at the University of Agriculture Faisalabad, combining rigorous academic theory with hands-on full-stack development experience. My primary focus is building engineering workflows that integrate artificial intelligence directly into SaaS platforms.
Instead of building generic wrappers, I focus on solving hard engineering challenges—like extracting blob-based assets from high-security advertising platforms, running stable automation engines in restricted cloud containers, and extracting chronological video frame sequences for visual reasoning.
I believe in modern development methodologies, where AI serves as a powerful accelerator for drafting boilerplate and structural templates, allowing developers to focus on architectural system design, edge cases, and runtime stability.
Experienced with
AI Integration & Agents
Deploying multimodal models, designing structured prompting architectures, and automating reasoning tasks.
Computer Vision Workflows
Chronological video/image parsing using OpenCV, frame sequence rendering, and object classification.
Automation & Scraping
Configuring headless web scrapers using Selenium and undetected-chromedriver to bypass strict ad verification walls.
Full-Stack Development
Designing fast async endpoints with FastAPI and building responsive user interfaces with React and TypeScript.
Containerized Deployment
Optimizing memory-management configs for Google Chrome inside Docker containers for reliable resource allocation.
AI-Assisted Workflows
Utilizing modern AI co-pilots to optimize planning, document architecture details, and accelerate implementation.
Technical expertise built on real-world implementation.
Frontend Development
- React
- TypeScript
- JavaScript
- Tailwind CSS
- Vite / Next.js
Backend Engineering
- Python
- FastAPI
- Node.js
- Express.js
- JWT Authentication
Artificial Intelligence
- Gemini API
- OpenRouter
- Prompt Engineering
- Machine Learning
- Deep Learning
- AI Agents
Computer Vision
- OpenCV
- Face Recognition
Automation Workflows
- Selenium
- Browser Automation
- Web Scraping
Infrastructure & Tools
- Docker
- Git / GitHub
- Render / Vercel Deployment
- SQLAlchemy ORM / Prisma
Advantage AI
Digital Marketing Creative Auditing Platform
The Problem
Performance marketing agencies spend significant resources testing ad creatives after launch, only learning what works from live campaign data. Advantage AI moves that evaluation earlier — auditing image and video advertisements before launch to identify weaknesses in hook effectiveness, visual pacing, copywriting quality, CTA placement, cognitive trigger alignment, and engagement potential.
Key Engineering Challenges Solved
High-Fidelity Video Analysis
Implemented chronological frame extraction using OpenCV and evaluated the full sequence within a single multimodal AI prompt — preserving narrative flow and significantly improving analysis accuracy over per-frame approaches.
Meta Creative Extraction
Developed Selenium automation using undetected-chromedriver to access blob-based video assets and programmatically bypass Meta's platform limitations that prevent direct media download.
Low-Resource Deployment
Optimized Chrome execution inside Docker containers with targeted memory management techniques to achieve stable deployment on Render free-tier infrastructure without performance degradation.
Technical Stack
Lessons Learned
- ✓AI-assisted development workflows
- ✓Full-stack architecture design
- ✓Production deployment challenges
- ✓Performance optimization
- ✓Multimodal AI integration
Planned Improvements
- ➔PostgreSQL + pgvector for semantic search
- ➔Celery distributed task processing
- ➔Multi-agent analysis architecture
- ➔Historical creative performance benchmarks
AI as a Development Collaborator
Acceleration, Not Delegation
Using AI to rapidly draft initial structures, write comprehensive unit test suites, and generate documentation—while maintaining complete human ownership over core logic and domain architecture.
System Design Priority
Recognizing that generating raw code is easy, but structuring state machines, defining strict TypeScript contracts, and designing resilient error boundaries requires deep human engineering.
Production-First Mindset
Building systems designed for real-world deployment constraints from day one—managing container memory footprints, handling API rate limits gracefully, and handling async task queues.
Academic Background
Bachelor of Science in Computer Science
University of Agriculture Faisalabad (UAF)
Focusing on core computer science foundations, algorithm complexity, software architecture, machine learning principles, database management systems, and object-oriented programming.