Margin Protector AI

Autonomous profit recovery for Shopify merchants.

Live Demo · Bazaar Demo · Report Bug

---

The Problem

Shopify merchants obsess over ad spend and conversion rates. Meanwhile, the real margin killer sits quietly in their warehouse: dead capital.

100 units of a $45 SKU with zero sales is $4,500 in trapped cash — depreciating at 2% per month in holding costs. Multiply that across 8 SKUs and you're bleeding $800+/month before a single customer bounces. By the time a human spots it, the damage is done.

Nobody is watching the back office. Until now.

---

The Solution

Margin Protector AI connects to your live Shopify store and runs 6 deterministic heuristic models against every product, order, and analytics data point — in under 50ms, on the edge, with zero LLM dependencies.

• Deterministic Stale Inventory Detection — Flags SKUs with 50+ units and zero velocity in 30 days. Calculates exact per-unit holding cost burn rate ($0.90/unit/month on a $45 SKU) and proposes clearance bundles with break-even timelines.

• Real-time Holding Cost Math — Every recommendation shows its work. Not "discount this" — instead: "Holding cost = $4,500 capital x 2% monthly = $90/mo. A 35% discount clears in ~14 days, recovering $2,925 in liquid capital. Break-even in 3 days."

• 1-Click Shopify GraphQL Execution — Each action card includes the exact discountCodeBasicCreate or purchaseOrderCreate mutation. Click "Deploy Fix" → preview the payload → "Active in Shopify" in 2 seconds.

• Shipping Margin Erosion Alerts — Flags products where shipping cost exceeds 40% of unit price. Models annual margin loss and proposes price adjustments with A/B test conversion impact projections.

• Conversion Dip Recovery — Detects days where conversion drops below 3.0%, estimates lost revenue using session data, and triggers flash sale campaigns with ROAS projections.

• Transparent AI Reasoning — Every action card has a collapsible "AI Thinking" panel showing exact dollar amounts, ROI percentages, and break-even analysis. No black boxes.

---

Why This Makes Money

The business model is simple: we recover money that's already being lost.

Metric	Before	After Margin Protector
Dead stock capital trapped	$12,000+	Recovered in 14 days
Monthly holding cost bleed	$800+/mo	$0 (stock cleared)
Time to identify leakage	2-4 weeks (manual)	< 50ms (automated)
Time to deploy Shopify fix	30+ min (manual)	1 click (2 seconds)

A store owner paying $99/mo for this tool recovers $800+ in the first month from dead stock alone. That's an 8:1 ROI before touching shipping optimization, conversion recovery, or refund analysis.

The TAM is every Shopify merchant with inventory — 4.6M+ stores.

---

Quick Start

# 1. Clone
git clone https://github.com/operatoruplift/Margin-Protector-AI.git
cd Margin-Protector-AI

# 2. Install
pnpm install

# 3. Run
pnpm dev

Environment Variables

Create .env.local in the project root:

SHOPIFY_ACCESS_TOKEN=shpua_your_token_here
SHOPIFY_STORE_URL=your-store.myshopify.com

---

Architecture

flowchart LR
    subgraph Client["Next.js 16 Dashboard"]
        UI["React 19 UI"]
        KPI["KPI Ribbon"]
        Terminal["AI Terminal"]
    end

    subgraph Edge["Vercel Edge"]
        API["/api/margin/analyze"]
        Discount["/api/shopify/discount"]
    end

    subgraph Engine["Margin Analyzer Engine"]
        H1["Dead Stock Detection"]
        H2["Shipping Erosion"]
        H3["Critical Restock"]
        H4["Conversion Dip"]
        H5["Refund Anomaly"]
        H6["Pending Backlog"]
    end

    subgraph Shopify["Shopify Admin API"]
        GQL["GraphQL Mutations"]
        Store["Store 30"]
    end

    UI --> API
    API --> Engine
    Engine --> API
    Terminal --> Discount
    Discount --> GQL
    GQL --> Store

    style Client fill:#18181B,stroke:#AAFF00,color:#fff
    style Edge fill:#09090B,stroke:#27272A,color:#fff
    style Engine fill:#09090B,stroke:#AAFF00,color:#fff
    style Shopify fill:#18181B,stroke:#7AB55C,color:#fff

Loading

We integrated Store 30's live data. To ensure zero latency and 100% uptime for this demo, the AI logic runs deterministically on the edge, evaluating live inventory counts against our margin-recovery heuristics. It instantly flags dead capital and queues up the exact Shopify GraphQL mutation needed to fix it.

---

The 6 Heuristics

#	Heuristic	Trigger	Urgency
1	Dead Stock Bundle	Inventory > 50, 0 sales / 30d	HIGH
2	Shipping Margin Erosion	Ship cost > 40% of price	HIGH
3	Critical Restock	Below reorder threshold	HIGH
4	Conversion Dip Recovery	CVR < 3.0% daily	MEDIUM
5	Refund Rate Anomaly	Refund rate > 5%	MEDIUM
6	Pending Order Backlog	4+ orders stuck pending	LOW

---

Tech Stack

Layer	Technology
Framework	Next.js 16.2.1 (App Router, Turbopack)
Runtime	React 19, TypeScript 5 (strict)
Styling	Tailwind CSS 4, shadcn/ui
Animation	Framer Motion 12, Magic UI
Commerce	Shopify Admin API (GraphQL)
Deploy	Vercel (Edge Functions)
AI Engine	Deterministic heuristics (edge-computed, <50ms)

---

Project Structure

src/
├── app/
│   ├── api/
│   │   ├── agent/run/        # AI analysis endpoint
│   │   ├── margin/analyze/   # Margin heuristics endpoint
│   │   ├── products/         # Product data
│   │   ├── orders/           # Order data
│   │   └── analytics/        # Analytics data
│   ├── page.tsx              # Dashboard + v2 pages
│   ├── layout.tsx            # Root layout
│   └── globals.css           # Design system + print styles
├── components/ui/            # shadcn/ui + Magic UI
├── lib/
│   ├── marginAnalyzer.ts     # 6-heuristic engine
│   ├── mockShopifyData.ts    # Store data layer
│   └── utils.ts              # Utilities
└── types/
    └── index.ts              # Shared interfaces

---

Team

Built by Operator Uplift at the OpenClaw Hackathon 2026.

License

MIT