DESIGN DECISION
AI assists; documents post
The assistant can query stock, prices, and history or prepare receiving and dispatch drafts, but it never bypasses human confirmation to modify inventory.
Vue 3Spring BootMyBatis-PlusMySQLRedisDocker
Every stock change stays traceable, and a sentence can become a reviewable business document.
About the Project
Built on RuoYi-Vue-Plus and RuoYi-WMS, this is my continuously evolving warehouse operations system for real small-business workflows. Products, storage locations, warehouses, customers, and suppliers are connected through receiving, dispatch, transfer, and stocktake documents, so inventory changes retain their business context.
The latest version includes an embedded AI assistant for natural-language inventory, price, and order-history queries. It can also turn a purchasing or sales request into a receiving or dispatch draft, while keeping a human confirmation step before the document enters the operational workflow. Unified search, product tags, stock alerts and sorting, customer-specific price history, a draggable 3D shelf view, warehouse and partner maps, and mobile-first actions support day-to-day work.
For delivery and operations, the project includes a Windows one-click launcher and desktop shortcut flow, Docker Compose production packaging, scheduled and manual database backups, legacy-schema alignment, and staged merge imports with duplicate detection, user mapping, and explicit conflict resolution.
AI-ASSISTED WAREHOUSE OPS
From one business request to a traceable stock change
AI interprets queries and prepares reviewable drafts; formal documents change inventory while maps, shelving, mobile tools, and operations share one data model.
4core stock flows
AIqueries + drafts
2desktop + mobile
01 ASK Natural language →
02 DRAFT Query / document →
03 REVIEW Human confirmation →
04 EXECUTE Stock document →
05 TRACE Ledger + history
Vue 3
Spring Boot
MySQL
Redis
COREWMS
CAPABILITIES
Operations and delivery loop
01Receive / dispatch / move / count✓
02AI queries and order drafts✓
03Maps and 3D shelving✓
04Backup and merge import✓
WAREHOUSE OPERATING SYSTEM
More than stock records: the way a warehouse works every day
Master data, business documents, AI assistance, spatial views, and data operations all revolve around one inventory truth.
4stock document types
5AI business tools
3delivery modes
CORE 01 · DOCUMENT-DRIVEN INVENTORY
Four documents create one traceable stock chain
Quantity is never treated as an isolated field. Every increase, decrease, location change, or reconciliation keeps a business source.
01 Purchase / receive
Select supplier, product, and bin; receiving details add inventory.
+ STOCK 02 Sale / dispatch
Use customer context and historical price, then deduct confirmed items.
− STOCK 03 Internal transfer
Record source and destination bins while total quantity stays unchanged.
MOVE 04 Stocktake
Reconcile recorded stock with physical counts and preserve the variance history.
RECONCILE Unified inventory ledger Product · SKU · warehouse · bin · quantity · operator · time
AI ASSISTANT
Ask in one sentence or prepare a reviewable document
The agent reads system data through controlled business tools rather than writing the database. Results can become receiving or dispatch drafts, but users still confirm the final document.
1
“Check PU tube inventory”
2
“What was the latest price for customer A?”
3
“Sell 10 meters of air tube to customer A”
4
“Receive 20 units of PU tube”
01
SEARCH SKUMatch names, specifications, and units
✓02
QUERY STOCKRead warehouses, bins, and availability
✓03
QUERY HISTORYRead customer dispatch and price history
✓04
RECEIPT DRAFTPrepare an editable receiving draft
✓05
SHIPMENT DRAFTPrepare an editable dispatch draft
✓ AI never bypasses permissions, posts stock directly, or skips human confirmation.
Per-user historyStreaming repliesVoice input
CORE 02 · OPERATIONAL VIEWS
One data model for different working contexts
Products, brands, and tags
Products carry brands, specifications, units, and batch tags; unified search spans master data and business documents.
Batch tagsUnit matchingUnified search
Statistics and alerts
Filter by warehouse, product, and tag; sort quantities and quickly isolate out-of-stock or low-stock items.
Tag filtersQuantity sortStock alerts
3D shelving and maps
Shelves can be dragged, scaled, rotated, and stored by warehouse; maps connect warehouses and business partners.
3D layoutMap pickerPartner distribution
On-site mobile flow
Price lookup, fast receiving, fast dispatch, and the AI assistant stay accessible with fewer interaction layers.
Price lookupFast documentsAI assistant
DATA & DELIVERY
Ready to deploy, migrate, recover, and keep upgrading
A business system matters beyond feature completion: legacy data must move, failures must be recoverable, and each environment must start reliably.
01 / BACKUP
Database backup
Create backups in the admin UI; a Docker task can run scheduled mysqldump + gzip jobs and retain records.
02 / MERGE IMPORT
Previewed merge import
Load into a temporary database, detect duplicates and conflicts, then let users choose mappings and resolution.
03 / SCHEMA ALIGN
Legacy schema alignment
Add missing tables, fields, and menus to existing deployments instead of rebuilding the database.
04 / DEPLOY
Multiple delivery paths
Windows one-click launch and desktop shortcuts coexist with Nginx, Spring Boot, Redis, and Docker Compose.
Skip duplicates→ Review conflicts→ Map user relations
SYSTEM ARCHITECTURE
A complete stack from interface to data operations
01 EXPERIENCE Vue 3 + Element Plus
Desktop console · mobile UI · AI chat · maps and 3D views
02 BUSINESS Spring Boot 3 + MyBatis-Plus
Documents · ledger · RBAC · AI orchestration · data operations
03 DATA MySQL + Redis + OSS
Business data · cache and sessions · attachments · backups
04 DELIVERY Nginx + Docker Compose
Static frontend · API proxy · backend · scheduled backup