Grzegorz Bogucki

URL - https://www.twine.net/signin

Executive Summary

A high-performance Direct-to-Consumer (D2C) e-commerce platform specifically engineered for the Trading Card Game (TCG) market. The project’s core innovation is Cardscan AI, a proprietary computer vision module that automates the entire supply chain: from physical card scanning and image recognition to real-time market valuation and automated logistics.

Key Technical & AI Achievements

• AI-Powered Inventory Automation (Cardscan): Developed a custom engine using Python (FastAPI) and Google Vision AI to handle OCR (Optical Character Recognition), image hashing, and identification of rare card variants. This module reduces product listing time from minutes to seconds.
• Dynamic Pricing & Real-time Data Intelligence: Implemented background Python scripts that integrate with global market APIs (Cardmarket, TCGPlayer) to synchronize prices automatically based on live market trends.
• Headless E-commerce Architecture: Leveraged a decoupled architecture using Medusa.js (Node.js) for the backend and Next.js (React) for the frontend. This ensures sub-second page loads and superior SEO.
• Full Logistics Automation: Built a custom integration with shipping aggregators. The system automatically generates shipping labels and tracking numbers upon purchase.
• Advanced DevOps: Orchestrated the environment using Docker, utilizing Nginx as a Reverse Proxy, PostgreSQL, and Redis for caching/task queuing.

Tech Stack

• AI & Backend: Python (FastAPI), Node.js (Medusa.js), Google Cloud Vision AI.
• Frontend: Next.js (SSR), React.js, TypeScript, Tailwind CSS.
• Infrastructure: Docker, PostgreSQL, Redis, Nginx.
• Integrations: Stripe (Payments), Furgonetka (Logistics).

Business Impact

This project solves the primary bottleneck in traditional e-commerce: manual data entry for unique, high-volume inventory. It is a SaaS-ready scalable solution designed to bridge the gap between complex AI algorithms and practical, revenue-generating business workflows.