Muhammad Ahmad

Senior Full-Stack AI Engineer & SaaS Architect

Advizi (Multi-Modal AI Content Engine) • Self-employed

Nov 2025 - Feb 2026 • 3 mos

Architected and engineered a production-grade Multi-Modal AI SaaS platform designed to transform product images into professional marketing assets (thumbnails,Product ads). This project serves as a high-performance benchmark for handling real-world revenue, complex AI orchestration, and high-concurrency scaling. Live Demo : https://advizi.vercel.app Key Engineering Achievements Atomic Financial Logic: Developed a high-integrity credit system using MongoDB atomic operators ($inc with pre-conditions). This prevents "double-spending" and fraud in high-concurrency environments without the overhead of distributed locks. Multi-Modal AI Orchestration: Integrated and orchestrated a suite of Google Gemini Models (3 Pro for images, VEO3 for video) into a unified content generation pipeline. Asynchronous Pipeline Design: Engineered a non-blocking, asynchronous video processing workflow. Utilizing a 202-Accepted pattern and client-side polling, the system handles 2-5 minute generation tasks without taxing server resources or hitting HTTP timeouts. Stateless Serverless Scaling: Implemented a stateless Vercel Serverless architecture, ensuring the backend can scale horizontally from 1 to 1,000+ concurrent requests automatically. Secure Webhook Integration: Built an idempotent webhook receiver for Clerk & Svix, ensuring user credits and subscription tiers are updated with 100% financial accuracy upon payment confirmation. Production-Grade DevOps: Integrated Sentry for real-time error tracking and Cloudinary CDN for optimized global asset delivery, reducing latency for end-users by up to 60%. Client/Business Insight Advizi was built with a "Production-First" mindset. By focusing on data consistency and stateless scalability, I delivered a platform capable of supporting a funded startup’s operational needs. The architecture minimizes infrastructure costs (Zero-Ops) while maximizing reliability for paying customers.

Full-Stack Developer | MERN E-Commerce Platform

Freelance Web Designer • Freelance

Aug 2025 - Dec 2025 • 4 mos

Designed and developed a production-ready full-stack e-commerce platform using the MERN stack with TypeScript. Built the complete system architecture—from schema design to frontend UI—prioritizing performance, security, and scalability. Live Demo : https://ecamp.vercel.app Key Responsibilities Full-Stack Architecture: Architected the entire application from scratch using MongoDB, Express, React, and Node.js (TypeScript). API Development: Designed and documented 60+ RESTful APIs utilizing a clean controller-service-repository pattern for modularity. Security & Auth: Implemented secure JWT-based authentication (access/refresh tokens) and Firebase Google Auth for seamless user onboarding. Transactional Systems: Developed complex logic for real-time cart management, multi-step checkout, and automated order processing. Payment Integration: Integrated Stripe API for secure transactions alongside a flexible Cash on Delivery (CoD) workflow. Admin Orchestration: Created a comprehensive Role-Based Access Control (RBAC) dashboard for managing inventory, orders, and business analytics. Optimization: Optimized MongoDB queries and implemented custom middleware for global error handling and data validation. Key Achievements Scalability: Designed a backend structure capable of handling concurrent requests through efficient load management and optimized DB indexing. Performance: Reduced API response times by optimizing database queries and implementing efficient data fetching patterns. Code Quality: Established a maintainable, type-safe codebase, significantly reducing runtime errors and improving development velocity. Impact: This project demonstrates my ability to deliver production-grade systems. By focusing on data consistency and security, I built a scalable solution capable of supporting the operational needs of a growing startup or retail business.