AI-powered automotive e-commerce platform transforming industrial product catalogs into sales-ready digital commerce operations.
MAD is a digital commerce platform developed for an industrial automotive lighting manufacturer. It combines an automotive webshop, B2B ordering flows, catalog operations, AI-assisted image processing, and automated publishing workflows.
This repository is a public portfolio showcase. Production source code, customer data, credentials, AI prompts, scraping logic, and deployment internals are intentionally private.
Organization context: MAD-AUTOMOTIVES represents the manufacturer identity behind this work. The public organization currently acts as the brand presence, while private source repositories remain access-controlled.
- Overview
- Business Problem
- Solution
- Key Features
- Product Screenshots
- Architecture Overview
- Technology Stack
- Future Roadmap
- Contact
Automotive manufacturers manage thousands of lighting references across brands, models, years, categories, product images, client accounts, and regional sales channels. Manual catalog publishing slows sales teams down and makes it difficult to maintain a consistent digital storefront.
MAD turns product data and digital assets into a structured e-commerce experience for B2B and B2C buyers.
- Large automotive catalogs are difficult to normalize and publish.
- Product images often require cleanup, classification, resizing, and validation.
- B2B customers need fast search, accurate fitment context, and streamlined ordering.
- Internal teams need admin tools without exposing operational complexity to customers.
MAD provides a commerce layer for automotive lighting manufacturers:
- Product catalog management for headlights, tail lights, mirrors, fog lights, and accessories.
- B2B ordering workflow for professional buyers.
- AI-assisted asset processing for image quality, classification, and publishing readiness.
- Admin and employee tools for catalog review, upload, validation, and product publication.
- Cloud-ready infrastructure for deployment, backups, and monitoring.
- Automotive product catalog with brand, model, category, and product metadata.
- Public storefront and B2B shop experience.
- Role-based admin, employee, commercial, and customer workflows.
- Product image automation and quality improvement.
- AI-assisted asset classification.
- Automated product publishing pipeline.
- Digital asset management for product and category media.
- Deployment-ready architecture using containerized services.
The AI layer supports product operations at a high level:
- Product image enhancement and quality checks.
- Image classification for catalog organization.
- Asset consistency improvements for storefront presentation.
- Workflow automation for upload, review, and publishing.
Exact prompts, model orchestration, validation rules, and processing internals are proprietary and not included in this showcase.
Screenshots should be exported from a sanitized demo environment only. Once captured, place them in screenshots/ using the names below so the repository reads like a finished product showcase.
| View | Purpose | File |
|---|---|---|
| Public Storefront | Shows the commerce experience | screenshots/01-hero-storefront.png |
| Operations Dashboard | Shows the internal workflow | screenshots/02-admin-dashboard.png |
| Product Catalog | Shows automotive catalog structure | screenshots/03-product-catalog.png |
| AI Asset Processing | Shows the AI product-image workflow | screenshots/04-ai-asset-processing.png |
| B2B Ordering | Shows business buyer workflow | screenshots/05-b2b-ordering.png |
- Hero storefront screenshot:
screenshots/01-hero-storefront.png - Main dashboard screenshot:
screenshots/02-admin-dashboard.png - Product catalog screenshot:
screenshots/03-product-catalog.png - AI asset workflow screenshot:
screenshots/04-ai-asset-processing.png - B2B order workflow screenshot:
screenshots/05-b2b-ordering.png
See docs/screenshots/README.md for capture guidelines.
flowchart LR
Buyer["B2B / B2C Buyer"] --> Storefront["Commerce Storefront"]
Staff["Internal Teams"] --> Portal["Admin & Operations Portal"]
Storefront --> API["Commerce API"]
Portal --> API
API --> Catalog["Catalog & Order Services"]
API --> Assets["Digital Asset Service"]
Assets --> AI["AI Image Processing"]
Catalog --> Database["Production Database Layer"]
Assets --> Storage["Cloud Object Storage"]
API --> Jobs["Background Publishing Jobs"]
The public architecture describes system responsibilities only. Production code, database schema, deployment files, certificates, and automation scripts remain private.
| Layer | Technologies |
|---|---|
| Frontend | Next.js, React, TypeScript |
| Styling | Tailwind CSS, component-driven UI |
| Backend | API service architecture |
| Data | Relational database layer |
| AI | Computer vision and image-processing services |
| Storage | Cloud object storage |
| Infrastructure | Docker, reverse proxy, cloud deployment |
| Operations | Monitoring, backups, CI/CD workflows |
- Advanced vehicle-fitment search.
- Dealer-specific price books and stock rules.
- More automated catalog enrichment.
- Analytics for sales, product demand, and customer behavior.
- Expanded AI quality scoring for product assets.
Recommended GitHub topics:
ecommerce, automotive, ai, nextjs, digital-commerce, automation
This project is available as a private technical walkthrough, architecture discussion, or client demo on request.