ZN SYSTEMS
Spare Parts AI Assistant with MCP and Claude
An AI-powered assistant that lets staff ask plain-language questions about customers, products, stock, suppliers, prices, and vehicle-specific parts.
Key results
Outstanding balances became easier to find
Customer recovery work became easier to target
Ledger checks moved into one assistant

Summary
We connected the shop database to Claude through a custom MCP server and built an assistant that understands plain-language questions. Staff can ask for customer history, purchasing trends, product details, prices, stock levels, supplier information, and vehicle-specific parts without needing database knowledge.
Context
Motorbike parts business
Project type
AI Automation
Primary stack
Claude
Focus areas
Technology
The situation
Important customer and ledger details were buried in business data.

The business was operating heavily on credit, which created serious cash-flow pressure. Important details were spread across customer records and ledgers, so it was difficult to see who owed money, how much they had purchased, how regularly they paid, and where balances were being missed.
The result
Outstanding balances became easier to find
The assistant helped the business find overlooked outstanding balances, group customers by spending and credit behaviour, and focus recovery work where it mattered. Staff could also check customer and manufacturer ledgers in one place instead of relying on a separate manual process, reducing effort and helping stop ongoing cash leakage.
Additional outcomes
- Customer recovery work became easier to target
- Ledger checks moved into one assistant
What was built
The main workflows and capabilities delivered for this project.
Plain-language questions over business data
Customer credit and payment history
Purchase trends and outstanding balances
Product, price, and stock lookup
Supplier and vehicle-specific parts lookup
Customer and manufacturer ledger checks
Have a software problem to solve?
Tell us what you are trying to build, improve, or automate. We can work out the right approach together.
Start a projectRelated cases
More work
Offline Field Sales & Inventory App
A mobile-first field sales app built on an existing MSSQL system, with offline sync, live stock lookup, orders, receipts, and packing workflows.
HR System for Fingerprint Attendance and Payroll
An on-premises HR system for fingerprint attendance, live staff monitoring, payroll calculations, and employee self-service.
Company Website Rebuild with Headless Next.js
A full website rebuild that moved a company away from WordPress and improved frontend performance, structure, and deployment.