ncreighton InventoryGenie — AI-Powered Inventory Management for Retailers
Overview
InventoryGenie is a production-ready inventory management system designed specifically for retail operations. It combines machine learning-powered demand forecasting, automated supplier integration, and real-time analytics to eliminate stockouts, reduce overstock, and optimize cash flow.
Why This Matters:
- Reduce Carrying Costs: Predict demand 30+ days ahead to avoid excess inventory
- Prevent Stockouts: Automated reordering keeps popular items in stock
- Multi-Channel Sync: Integrates with Shopify, WooCommerce, Square, and custom APIs
- Real-Time Visibility: Dashboard shows stock levels, turnover rates, and supplier performance across all locations
- Supplier Automation: Auto-generates POs, tracks delivery times, and flags slow vendors
Key Integrations:
- Shopify (inventory sync, order data)
- WooCommerce (product catalog, sales history)
- Slack (low-stock alerts, reorder notifications)
- Google Sheets (custom reporting, stakeholder dashboards)
- Stripe (sales velocity correlation)
- QuickBooks (cost accounting)
Quick Start
Try these prompts immediately to see InventoryGenie in action:
Example 1: Demand Forecast Report
Generate a 60-day demand forecast for my top 20 SKUs.
Include seasonality adjustments, confidence intervals, and recommended reorder quantities.
Use the past 18 months of sales data from Shopify.
What you'll get: CSV with SKU, predicted demand, safety stock, reorder point, and optimal order quantity.
Example 2: Automated Low-Stock Alert
Set up automated alerts for any product that falls below 30% of average weekly sales.
When triggered, create a Slack message to #inventory-team and auto-generate a draft PO to our main supplier.
What you'll get: Real-time monitoring + Slack notifications + supplier PO drafts ready for approval.
Example 3: Supplier Performance Dashboard
Analyze supplier performance for the last 90 days.
Show on-time delivery rate, average lead time, cost per unit trend, and quality issues.
Recommend which suppliers to prioritize based on reliability.
What you'll get: Ranked supplier scorecard with actionable recommendations.
Example 4: Multi-Location Inventory Optimization
I have 5 retail locations. Show me which products are overstock in one location
but understocked in another. Suggest transfers to minimize safety stock across the network.
What you'll get: Transfer recommendations with cost savings estimates.
Capabilities
1. AI-Powered Demand Forecasting
- Models Supported: Prophet (seasonal), ARIMA, XGBoost, Linear Regression
- Data Sources: Shopify, WooCommerce, custom CSV uploads
- Forecast Horizon: 7 to 180 days ahead
- Accuracy Metrics: MAPE, RMSE, confidence intervals (80%, 95%, 99%)
- Seasonal Adjustments: Auto-detects holidays, promotions, and cyclical patterns
- Multi-Variable Support: Correlates with weather, marketing spend, competitor activity
Usage Example:
Forecast demand for "Winter Boots" using Prophet model.
Include promotional impact from our Black Friday campaign.
Show 95% confidence interval for safety stock calculation.
2. Automated Supplier Integration
- Supported Suppliers: Shopify, custom APIs, EDI files, email-based ordering
- PO Generation: Auto-creates purchase orders with optimal quantities
- Lead Time Tracking: Monitors supplier delivery performance
- Cost Optimization: Suggests bulk order discounts and consolidation opportunities
- Compliance: Validates MOQ (minimum order quantity), packaging, incoterms
Usage Example:
Connect to our supplier API (endpoint: api.supplier.com/v2/orders).
When reorder point is triggered, auto-generate a PO for 2 weeks' worth of inventory.
Set lead time buffer to 5 days and include 10% safety stock.
3. Real-Time Analytics & Reporting
- Dashboard Metrics: Stock turnover, carrying cost, stockout frequency, supplier KPIs
- Custom Reports: Export to Google Sheets, PDF, or Slack
- Alerts & Thresholds: Configurable triggers for low stock, overstock, slow movers
- Inventory Aging: Identifies obsolete stock for clearance
- ABC Analysis: Categorizes SKUs by sales velocity and profitability
Usage Example:
Create a daily Slack report showing:
- Top 10 fast-movers (risk of stockout)
- Top 10 slow-movers (excess inventory)
- Suppliers with delayed shipments
- Forecast accuracy vs. actual sales
4. Multi-Location Management
- Network Optimization: Minimizes total safety stock across locations
- Transfer Recommendations: Suggests inventory redistribution
- Centralized Visibility: Single dashboard for all warehouses and retail stores
- Demand Pooling: Shares stock buffers across locations to reduce redundancy
Usage Example:
Analyze inventory across 5 locations.
Show me products that are overstock in Location A but understocked in Location B.
Calculate transfer costs and recommend optimal moves.
5. Inventory Aging & Obsolescence Detection
- Days on Hand (DOH): Tracks how long inventory sits
- Turnover Velocity: Identifies slow-moving SKUs
- Markdown Recommendations: Suggests clearance pricing for stale stock
- Write-off Analysis: Flags items for potential loss
Configuration
Required Environment Variables
# Shopify Integration
export SHOPIFY_API_KEY="your_shopify_api_key"
export SHOPIFY_API_PASSWORD="your_shopify_password"
export SHOPIFY_STORE_URL="your-store.myshopify.com"
# Forecasting
export FORECAST_MODEL_TYPE="prophet" # Options: prophet, arima, xgboost, linear
export FORECAST_HORIZON_DAYS="60"
export CONFIDENCE_INTERVAL="95"
# Supplier Integration
export SUPPLIER_API_ENDPOINT="https://api.supplier.com/v2"
export SUPPLIER_API_KEY="your_supplier_key"
# Alerts & Notifications
export SLACK_WEBHOOK_URL="https://hooks.slack.com/services/YOUR/WEBHOOK/URL"
export LOW_STOCK_THRESHOLD_PERCENT="30" # Trigger alert at 30% of avg weekly sales
# Database
export INVENTORY_DB_HOST="localhost"
export INVENTORY_DB_PORT="5432"
export INVENTORY_DB_NAME="inventory_prod"
Setup Instructions
Connect Data Source
# For Shopify inventorygenie connect shopify --api-key $SHOPIFY_API_KEY --password $SHOPIFY_API_PASSWORD # For WooCommerce inventorygenie connect woocommerce --url https://yourstore.com --api-key $WC_KEYConfigure Forecasting
inventorygenie config forecast \ --model prophet \ --horizon 60 \ --confidence 95 \ --seasonality autoSet Up Alerts
inventorygenie alert add \ --name "low_stock" \ --condition "stock < (avg_weekly_sales * 0.3)" \ --action "slack" \ --webhook $SLACK_WEBHOOK_URLEnable Supplier Automation
inventorygenie supplier connect \ --name "primary_supplier" \ --endpoint $SUPPLIER_API_ENDPOINT \ --auto-reorder true \ --lead-time 5
Example Outputs
Demand Forecast Report
SKU,Product Name,Avg Weekly Sales,Forecast 30d,Forecast 60d,Safety Stock,Reorder Point,Recommended Order Qty
SKU-001,Wireless Headphones,145,4350,8700,290,435,1450
SKU-002,USB-C Cable,312,9360,18720,625,936,3120
SKU-003,Phone Case,87,2610,5220,174,261,870
Low-Stock Alert (Slack Format)
🚨 LOW STOCK ALERT
Product: Wireless Headphones (SKU-001)
Current Stock: 125 units
Reorder Point: 435 units
Forecast Lead Time: 5 days
Recommended Order: 1,450 units
Estimated Cost: $14,500
Action: Auto-PO sent to Primary Supplier ✓
Supplier Performance Scorecard
Supplier Name: Primary Supplier
On-Time Delivery Rate: 94.2% ✓
Avg Lead Time: 4.8 days (target: 5)
Cost per Unit Trend: -2.3% (improving) ✓
Quality Issues: 2 in 90 days (0.8%) ✓
Overall Score: A+ (Recommended for priority)
Multi-Location Optimization
Transfer Recommendation:
FROM: Location A (Overstock)
TO: Location B (Understock)
Product: Winter Boots (SKU-042)
Quantity: 45 units
Current Stock Location A: 180 units (45 days supply)
Current Stock Location B: 8 units (2 days supply)
Transfer Cost: $180
Savings: $1,200 (avoided safety stock)
Tips & Best Practices
1. Optimize Forecast Accuracy
- Feed Historical Data: Provide at least 12 months of sales history for best results
- Exclude Anomalies: Remove one-time bulk orders or promotional spikes from training data
- Update Seasonality: Adjust seasonal factors quarterly as your business evolves
- Monitor MAPE: Aim for <15% mean absolute percentage error; if higher, review data quality
2. Supplier Relationship Management
- Lead Time Buffer: Add 20-30% buffer to supplier lead times for safety
- Diversify Suppliers: Use secondary suppliers for critical SKUs to avoid single-point failures
- Track Performance: Review supplier scorecards monthly; switch if on-time delivery drops below 90%
- Negotiate Bulk Discounts: Use forecast data to negotiate better pricing for predictable volumes
3. Inventory Segmentation (ABC Analysis)
- A Items (High value, high velocity): Reorder frequently, tight safety stock
- B Items (Medium): Standard reorder points
- C Items (Low value): Higher safety stock, less frequent reorders
- Apply Different Policies: Use tighter forecasts for A items; looser for C items
4. Reduce Carrying Costs
- Target Turnover Ratio: Aim for 4-8x annual turnover depending on category
- Quarterly Reviews: Identify slow movers; mark down or clear stock
- Just-In-Time for High-Volume: Reduce safety stock for fast-moving items with reliable suppliers
- Seasonal Planning: Build inventory 60 days before peak season; clear 30 days after
5. Multi-Location Strategy
- Centralized Safety Stock: Keep buffer stock at warehouse; distribute to stores as needed
- Cross-Location Transfers: Use forecasts to identify and execute transfers before stockouts
- Network Demand Pooling: Reduce total inventory 10-20% by sharing stock across locations
Safety & Guardrails
What This Skill Will NOT Do
⚠️ No Financial Decisions Without Review
- InventoryGenie generates recommendations, not binding decisions
- All POs over $5,000 require manual approval (configurable)
- Does not execute payments or change supplier contracts
⚠️ No Data Deletion or Modification
- This skill reads inventory data; it does not delete or permanently modify records
- All changes are logged and auditable
- Forecast adjustments are suggestions only
⚠️ No Supplier Communication Without Approval
- Auto-generated POs are drafts; they require human review before sending
- Slack alerts notify your team; they don't commit to orders
- Email to suppliers is always sent by your account (not InventoryGenie's)
⚠️ Forecast Accuracy Limitations
- Forecasts assume historical patterns continue
- Cannot predict sudden market disruptions, supply chain crises, or competitive changes
- Accuracy decreases beyond 90 days; use with caution for long-term planning
- Always validate forecasts with domain expertise
⚠️ Data Privacy & Security
- Requires read access to your inventory and sales data
- Does not share data with third parties
- Encrypts all API credentials in transit and at rest
- Complies with GDPR, CCPA, and SOC 2 standards
⚠️ Supplier Integration Limitations
- Only works with suppliers that provide API access or EDI files
- Email-based ordering requires manual PO entry
- Does not negotiate pricing or enforce contract terms
- Lead time estimates are historical averages; actual times may vary
Troubleshooting
Common Issues & Solutions
Q: My forecasts are inaccurate (MAPE > 25%)
A: Check these items:
- Is your sales data clean? Remove returns, cancellations, and bulk orders from training data
- Do you have enough history? Minimum 12 months; ideally 24 months
- Are you missing seasonality? Adjust seasonal factors for holidays and promotions
- Is your product new? New SKUs have no history; use similar products as proxy
- Try a different model: Switch from Prophet to XGBoost if you have >50 SKUs
Q: Supplier POs aren't being created
A: Verify:
- Supplier API credentials are correct:
inventorygenie test supplier --name primary_supplier - Lead time is set:
inventorygenie config supplier --lead-time 5 - Reorder point is being triggered: Check current stock vs. calculated reorder point
- Approval threshold: If PO value > $5,000, it requires manual approval (check email)
Q: Low-stock alerts aren't firing
A: Debug:
- Confirm Slack webhook is valid:
inventorygenie test alert slack - Check threshold: Is current stock actually below 30% of weekly average?
- Review alert rules:
inventorygenie alert list - Check logs:
inventorygenie logs --filter alert --last 24h
Q: Forecast accuracy drops after promotions
A: This is expected. Solutions:
- Exclude promotional periods from training data (mark as anomalies)
- Increase safety stock during promotional windows
- Use separate forecasts for "baseline" vs. "promotional" demand
- Update forecasts daily during campaigns to catch real-time trends
Q: Multi-location transfers are too expensive
A: Optimize:
- Batch transfers: Consolidate 5+ SKUs per shipment
- Adjust transfer cost model:
inventorygenie config transfer --cost-per-unit 2.50 - Increase transfer threshold: Only recommend transfers for items with >$1,000 savings
- Use slower shipping: Trade cost for longer lead time
Q: API rate limits exceeded
A: Solutions:
- Reduce forecast frequency: Change from hourly to daily updates
- Batch API calls: Consolidate multiple requests into single endpoint call
- Use webhook polling instead of constant API hits
- Contact your data source (Shopify, WooCommerce) for higher rate limits
FAQ
Q: How often should I update forecasts? A: Daily for fast-moving items (turnover > 2x/week), weekly for standard items, monthly for slow movers. More frequent updates = better accuracy but higher compute cost.
Q: Can I use this with multiple sales channels? A: Yes. InventoryGenie aggregates data from Shopify, WooCommerce, Amazon, and custom APIs. It provides unified forecasting across all