All Projects
Meguri
SmartStock for recipe-based businesses and small retail
Overview
Meguri is a smart stock and demand forecasting app for recipe-driven businesses and small retail teams. It keeps workspace, inventory, recipes, sales, suppliers, purchase planning, alerts, and forecasting in one operational system so daily work stays visible and controlled.
Problem & Solution
The Problem
Small businesses often manage stock in spreadsheets or manual workflows. That makes stock movement hard to track, connects sales to recipes or BOMs poorly, delays purchasing, and increases the chance of stockouts or waste.
The Solution
Meguri solves this with a workspace-scoped app built on Next.js and Convex, with real authentication, inventory CRUD, stock movement audit logs, sales tied to recipes, supplier and purchase planning, alerts, and forecasting or reorder recommendations based on live system data.
Tech Stack
Provides the app shell and interactive UI layer for the main workspace, dashboard, inventory, forecasting, alerts, and planning pages.
Keeps the data model, query handlers, and UI contracts explicit across the repo, which matters for workspace-scoped business logic.
Acts as the backend, database, and realtime query layer for inventory, sales, supplier data, alerts, and forecasting snapshots.
Handles authentication and session management so workspace access control can be enforced reliably.
Provides the styling system for the dashboard, navigation, and operational pages without adding heavy custom CSS overhead.
Supply the UI motion and icon set used to keep the experience clear, modern, and easy to scan.
Key Features
- Multi-tenant workspace model with role-based access control
- Inventory management with add, edit, archive, adjustment, and movement history
- Product and recipe/BOM mapping to connect finished goods with ingredients
- Sales transactions that deduct stock from recipes and preserve an audit trail
- Supplier management and purchase planning through reorder recommendations
- Forecasting dashboard with 7, 14, and 30 day snapshots plus manual refresh
- Alerts for low stock and unusual demand
- Dashboard pages for inventory, products, sales, forecasting, suppliers, reports, and settings
Challenges & Learnings
Workspace scoping and permissions
Every query and mutation has to stay inside the correct tenant boundary so data cannot leak across workspaces or bypass role-based permissions.
Forecasting and recommendation quality
The current forecasting layer uses moving averages from stock movement data, which works for the MVP but will need stronger forecasting logic over time.
Idempotent reorder generation
Purchase recommendations need to be regenerated safely without creating duplicate pending records, so existing recommendations are cleared before a new run.
Production readiness gaps
The roadmap still calls out work such as deployability baseline, supplier PO flow, reports and export, onboarding, and multi-location support before production hardening is complete.
Screenshots
Meguri dashboard overview
Meguri product creation modal
Meguri inventory list
Meguri recipe and BOM screen
Meguri sales log
Meguri forecasting view
Meguri supplier management screen