AI-Powered Shopping Assistants for Spryker using Model Context Protocol

Dmitriy Krainiy · June 10

Hello CommerceQuest community! 👋

I've been working on an exciting proof of concept that bridges AI technology with Spryker Commerce OS, and I'd love to get your thoughts and feedback on this approach.

What I Built

I've developed a Spryker MCP (Model Context Protocol) Server that enables AI assistants (like Claude, ChatGPT, or custom AI agents) to interact directly with Spryker APIs, creating shopping experiences through natural language.

Repository: https://github.com/spryker-sdk/spryker-mcp-server

What It Does (Current POC Features)

Instead of customers clicking through categories and filters, they can now shop like this:

"Find me wireless headphones under $200 with good reviews."
"Add this laptop to my cart and check shipping options."
"I need running shoes for trails, what do you recommend?"

The AI understands the request, searches products, manages carts, handles checkout, and provides customer service - all through natural conversation.

🎯 E-commerce Operations Covered:

Product Discovery & Search

Natural language product search ("Find me wireless headphones under $200")
Intelligent filtering, sorting, comparison, and recommendations
Real-time inventory and pricing information

Shopping Cart Management

Add/remove items with conversational commands
Cart optimization suggestions
Support for both guest and authenticated customers

Checkout & Orders

AI-guided checkout assistance
Payment and shipping option recommendations
Order tracking and history access for authenticated customers
Support for both guest and authenticated customers

Customer Experience

Customer authentication
Guest customers supported

⚙️ Technical Approach:

Built on Model Context Protocol
TypeScript with full type safety
Multiple transport methods (stdio, HTTP, SSE)

🤔 Questions for the Community

I'm particularly interested in your thoughts on:

What e-commerce scenarios would benefit most from conversational AI?
What concerns would you have deploying this in production?
Are your clients/customers asking for AI shopping experiences?
What capabilities would be most valuable to add next?

🚧 Current Limitations (POC Status)

This is a proof of concept, so please note:

Demo/development environment only - not production-ready
Limited error handling for edge cases
Basic security implementation - needs enterprise hardening
No advanced analytics or reporting features yet
Simplified configuration - needs more flexible setup options

Has anyone else been exploring AI integration with Spryker or other commerce platforms?

What approaches have you tried, and what lessons learned can you share?

---