Reduce High AI API Costs for LLM Applications

# Install Redis client
npm install ioredis

# Basic cache pattern
import Redis from "ioredis"
const redis = new Redis()

async function getCached(key: string, fetcher: () => Promise<unknown>) {
  const cached = await redis.get(key)
  if (cached) return JSON.parse(cached)
  const data = await fetcher()
  await redis.set(key, JSON.stringify(data), "EX", 300)
  return data
}

-- Add index on commonly queried columns
CREATE INDEX idx_orders_user_id ON orders(user_id);
CREATE INDEX idx_logs_created_at ON logs(created_at);

-- Check query execution plan
EXPLAIN ANALYZE SELECT * FROM orders WHERE user_id = $1;

# AWS — find unattached EBS volumes
aws ec2 describe-volumes \
  --filters Name=status,Values=available \
  --query 'Volumes[*].{ID:VolumeId,Size:Size}'

# GCP — list idle VMs
gcloud compute instances list \
  --filter="status=TERMINATED"

# AWS — create a budget alarm
aws budgets create-budget \
  --account-id 123456789012 \
  --budget file://budget.json \
  --notifications-with-subscribers file://notify.json

# Or use your cloud console's budget dashboard

const prompt = `You are a cloud diagnostics expert.

Given the following system issue, respond with:
1. Root cause (one sentence)
2. Fix steps (numbered list)
3. Prevention tips (bullet list)

Rules:
- Be specific and actionable
- Do not hallucinate services the user didn't mention
- If uncertain, say so explicitly

Issue: ${userInput}`

import { z } from "zod"

const AnalysisSchema = z.object({
  problem: z.string().min(10),
  cause: z.string().min(10),
  fix: z.array(z.string()).min(1),
  confidence: z.number().min(0).max(1),
})

const parsed = AnalysisSchema.safeParse(modelOutput)
if (!parsed.success) {
  console.error("Invalid output:", parsed.error.flatten())
}