Can I vibe code Peel Insights?

Build a standalone full-stack customer cohort and LTV analytics dashboard for Shopify using Next.js, TypeScript, Tailwind CSS, PostgreSQL, and Prisma ORM.

1. DATA MODEL:
Create database models for:
- `Store`: id, shopifyDomain, accessToken, status.
- `Customer`: id, storeId, shopifyCustomerId, firstOrderAt, totalSpent, ordersCount.
- `Order`: id, storeId, customerId, shopifyOrderId, grossAmount, netAmount, taxAmount, shippingAmount, currency, createdAt.
- `CohortMetric`: id, storeId, cohortMonth (YYYY-MM), monthNumber (0-12), totalCustomers, retainedCustomers, cumulativeRevenue, averageOrderValue.

2. DATA INGESTION & PROCESSING:
- Implement a Webhook handler (`/api/webhooks/shopify/orders-create`) that records incoming orders and upserts customer order counts, adjusting total spent and first purchase date.
- Implement a daily cron job that aggregates historical order data into cohort months based on a customer's first purchase date.
- Write SQL aggregation logic calculating cumulative LTV (Month 0, Month 1, Month 3, Month 6, Month 12) per acquisition cohort.

3. CORE UI & DASHBOARD:
- Render a standard interactive Cohort Retention Matrix (Rows = First Purchase Cohort Month, Columns = Month 0 to Month 12).
- Toggle cohort view between 'Retention Rate (%)', 'Cumulative Revenue ($)', and 'Average LTV per Customer ($)'.
- Color-code table cells dynamically using heat map shading (e.g., higher retention/LTV represented by darker green background shades).
- Include a high-level summary dashboard displaying: Overall 60-day Retention Rate, Average 90-day LTV, and RFM Customer Segments (Champions, At-Risk, Lost).

4. FAILURE HANDLING & CONSTRAINTS:
- Ensure order ingestion is idempotent based on `shopifyOrderId` to prevent double-counting.
- Handle multiple currencies by converting order amounts using standard fixed USD rate assumptions or Shopify line-item currency data.
- Out of scope: Live ad platform integrations (Meta/Google Ads attribution) and direct SMS/Email campaign execution.

$ each button prefixes agent-specific run instructions · build your own product, never copy proprietary code, trademarks or designs