AI Prompt for Storage Decisions

Decide where data lives with this AI prompt: database vs cache vs memory analysis on durability, latency, size, and operational cost trade-offs.

The prompt

Role: You are a systems architect who decides where data lives.

Context: Data: [DATA_DESCRIPTION]. Size/growth: [SIZE]. Access pattern: [ACCESS_PATTERN]. Latency target: [LATENCY]. Durability need: [DURABILITY: survive restarts / ephemeral OK / must never lose]. Budget/complexity tolerance: [TOLERANCE].

Task: 1) Evaluate the options: in-memory structure, embedded DB, cache (Redis/Memcached), primary database, or hybrid — against durability, latency, size, and operational complexity. 2) Recommend ONE (or a hybrid with clear responsibilities per layer). 3) Justify with numbers: can it hit [LATENCY] at [SIZE]? What happens on restart/failure? 4) Sketch the data flow (writes and reads through the layers). 5) List the operational costs (backups, monitoring, failover) so the choice is eyes-open.

Format: Option evaluation table, then recommendation, then the numbers justification, then data flow, then ops costs.

Ask the AI to sketch the failure drill: what you do when each layer fails, in order. Storage decisions are only as good as their disaster plans.

Constraints: Be honest about operational burden — a Redis cluster for ephemeral data is over-engineering. Match the solution to [TOLERANCE], not to ideal architecture.

Variables to fill in

Replace each [variable] in the prompt with your own details:

[DATA_DESCRIPTION] [SIZE] [ACCESS_PATTERN] [LATENCY] [DURABILITY] [TOLERANCE]

Example input

DATA=user sessions, SIZE=10M sessions/day, ACCESS=key lookup by session id, 1ms target, PERSISTENCE=survive restarts, STACK=Redis vs Postgres vs in-memory

How to use this prompt

State the durability requirement honestly — 'survive restarts' vs 'cache only' is the key fork. Include access latency targets; they rule out options fast.