Building SupplyShock Sentinel with MeDo: An AI Control Tower for Supply Chain Risk
Supply chains usually fail quietly before they fail visibly.
A supplier delay starts as a small update. Inventory drops below a safe level. A Tier-2 supplier gets disrupted. A purchase order moves from “confirmed” to “delayed.” By the time the business notices, the problem may already have turned into a customer delivery failure.
That is the problem I wanted to solve with SupplyShock Sentinel.
Live app: https://app-bfiq7qjaez9d.appmedo.com/
What is SupplyShock Sentinel?
SupplyShock Sentinel is an AI-powered supply-chain risk analysis and decision-support platform for small and mid-sized businesses.
It helps users answer one simple but important question:
What will fail first, why will it fail, and what should we do now?
The app allows users to upload supply-chain data, validate it, analyze it, simulate disruptions, and generate recovery actions.
Users can either:
- try built-in demo data
- upload their own CSV files
Supported data types include:
- Purchase Orders
- Suppliers
- Inventory
- Supplier Relationships
- Alerts
Why I built it
Many small and mid-sized businesses still manage supply-chain operations using Excel, email, WhatsApp, and manual follow-ups.
That creates major blind spots:
- supplier delays are discovered too late
- inventory stockouts are not predicted early
- supplier reliability is not quantified
- Tier-2 and Tier-3 dependencies are invisible
- teams do not know what to fix first
- financial exposure is not estimated before disruption
Most enterprise supply-chain control towers are expensive and complex. I wanted to build a lighter, accessible version that a smaller business could understand and use quickly.
How MeDo helped
I built the app using MeDo, an AI-powered app-building platform that helps generate full-stack web applications through natural language prompts.
MeDo made the process much faster because I did not need to manually build every page, route, form, table, and dashboard from scratch.
Instead, I could describe the app as a product:
- what pages it needed
- what data structures it should use
- what the upload flow should look like
- how risk scoring should work
- how dashboards should react
- how users should edit data
- how what-if scenarios should work
Then I could test, find gaps, and ask MeDo to improve specific parts.
The biggest benefit was iteration speed.
The app started as a dashboard idea. After testing, I realized that a dashboard alone was not enough. Users needed a clear upload workflow. So I used MeDo to add:
- CSV template downloads
- required header explanations
- upload preview
- row-level validation
- import controls
- reactive dashboard updates
- data editing
- what-if scenario testing
- Help and About pages
MeDo helped turn the project from a demo into a usable product experience.
Core workflow
The app follows this journey:
- Open the app.
- Choose Try Demo Data or Upload Your Own Data.
- Review required CSV formats.
- Download sample templates.
- Upload supply-chain data.
- Preview and validate uploaded rows.
- Confirm import.
- Run Sentinel Analysis.
- View dashboards and alerts.
- Edit data or run what-if scenarios.
- Ask AI Copilot what to fix first.
- Download risk-scored CSV files and reports.
This was important because a first-time visitor should not need external explanation to understand the app.
Risk scoring logic
The app uses an explainable risk scoring model.
text
Risk Score = Inventory Risk + Supplier Risk + Delay Risk + Priority Risk + Status Risk + Deep-Tier Risk
Top comments (0)