Data processing optimizations: eliminate N+1 queries and add performance indexes

[Copilot]

Multiple query inefficiencies were causing poor performance at scale: N+1 patterns in order creation, in-memory aggregations, missing pagination, and absent indexes on search/join columns.

Query Pattern Fixes

Order creation N+1 → batch operations

// Before: 1 + 2N queries
for (const item of items) {
  const prod = await db.select().from(products).where(eq(products.id, item.productId));
  await db.insert(orderItems).values({...});
}

// After: 3 queries total
const productsList = await db.select().from(products).where(inArray(products.id, productIds));
const productsMap = new Map(productsList.map(p => [p.id, p]));
await db.insert(orderItems).values(orderItemsValues);

AI model stats: in-memory filtering → SQL aggregation

• Replaced SELECT * + JS filter with COUNT() and conditional SUM()
• Eliminates dataset transfer and O(n) client-side processing

Wallet operations: eliminated redundant SELECTs

• Removed duplicate getWalletByUserId calls in update paths
• 4-5 queries → 2-3 queries per operation

Users count: conditional join

• Count query only joins roles table when role filter present

Pagination

Added to getOrders (limit: 20) and getAdminLogs (limit: 50) to prevent unbounded result sets.

Database Indexes

Generated migration 0010_little_exodus.sql with 19 indexes:

• Search fields: name, username
• Foreign keys: user_id, role_id, admin_id, order_id, product_id
• Sort fields: created_at
• Composite: (status, created_at), (user_id, status)

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.