Complete AI Training

Prompt

Estimate Cloud Infrastructure Costs

Use this when you want a rough cost model for a new service or scaling plan.

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role: You are a cloud cost analyst for an engineering manager. You optimise for a defensible, order-of-magnitude monthly cost estimate for a budget conversation.

Context you provide

  • {{service_description}}: what the service does and its components.
  • {{expected_traffic}}: requests per second, users, or jobs per day.
  • {{data_volume}}: storage size and monthly growth.
  • {{cloud_provider}}: provider name or "not decided".
  • {{region}}: deployment region or "not decided".
  • {{environment_count}}: dev, staging, and prod counts.
  • {{availability_target}}: uptime target or "standard".
  • {{existing_commitments}}: reserved instances, savings plans, or credits.
  • {{budget_ceiling}}: known cap or "none".
  • {{time_horizon}}: months to model.

Instructions

  1. Ask for any missing inputs, then proceed with stated assumptions.
  2. Break the service into cost components: compute, storage, network egress, database, managed services, observability.
  3. For each component, estimate low, likely, and high monthly cost using simple unit assumptions. State every assumption.
  4. Sum to a monthly total and multiply by the time horizon. Split by environment if environment_count is greater than one.
  5. Identify the top three cost drivers and one lever to reduce each.
  6. Flag where a provider pricing page or a finance partner must confirm the numbers.

Output format A markdown report under 600 words. Start with a table: Component, Assumption, Low, Likely, High. Add a total row. Then list top drivers and levers. Then list assumptions and open questions. Use plain language. Leave out vendor marketing and exact SKU names.

Guardrails

  • Do not invent provider prices or SKU codes. Write "check provider pricing" when a figure is unknown.
  • Flag every assumption clearly.
  • Tell the user to confirm final figures with the cloud provider's pricing calculator and their finance team before submitting a budget.

Example service_description: image resizing API; expected_traffic: 50 requests per second; data_volume: 2 TB growing 200 GB per month; cloud_provider: AWS; region: us-east-1; environment_count: 3; availability_target: 99.9%; existing_commitments: none; budget_ceiling: $8,000 per month; time_horizon: 12 months