Skip to main content

2.1 Functional and infrastructure requirements

Three layers, in order​

LayerQuestionExample
Business requirementWhat outcome does the organisation want?"Cut invoice-processing headcount cost by 40% this fiscal year"
Functional requirementWhat must the system do?"Extract 14 named fields from PDF invoices and post them to the ERP; flag any invoice where a field is missing or ambiguous"
Infrastructure requirementUnder what constraints must it run?"12,000 invoices/night, finished by 06:00, EU data residency, no PII in logs, $400/month ceiling"

Items in this skill hand you a business sentence and ask which requirement is missing, or which platform decision it forces. The trap is answering with a technology before the requirement is stated.

Requirement → platform decision​

Every infrastructure requirement maps onto a specific, nameable Claude decision.

RequirementDecision it forces
Results needed now, in a UISynchronous Messages API plus streaming so time-to-first-token is low
Results needed by tomorrow, in volumeMessage Batches API — 50% cost, within 24 h
Hard per-request cost ceilingModel tier plus effort plus prompt caching; measure with usage
Hard latency budgetSmaller model, lower effort, shorter output, streaming; fast mode on Opus 5 where available
Large stable context reused across requestsPrompt caching with the stable prefix first
Very long inputs (whole books, huge repos)A 1M-context model, plus chunking or context management
Machine-consumable output for a downstream systemStructured outputs plus schema validation before the write
Data residency in a regioninference_geo on the request, or a provider/region deployment
Procurement is locked to a cloud vendorClaude through Amazon Bedrock, Google Vertex AI, or Microsoft Foundry — see availability tradeoffs
Zero/short data retention requiredCheck per-model availability — some models are not offered under zero data retention
Auditability of every actionA workflow with logged steps, not an autonomous agent; plus structured provenance on outputs
Irreversible operations in scopeA deterministic gate or human approval, not a prompt instruction
Throughput above your rate limitBatch, request-level concurrency control with backoff, or a priority/committed-capacity arrangement

Non-functional requirements the guide expects you to elicit​

  • Volume and shape: requests per day, tokens per request, peak concurrency, input modality (text, images, PDFs).
  • Latency: p50 and p95 targets, and whether time-to-first-token or time-to-last-token is the one users feel.
  • Cost: per request, per document, per month — and who owns the budget.
  • Accuracy target: expressed as a measurable rate on a labelled set, not "it should be accurate".
  • Failure behaviour: what happens when the model is wrong, when the API is down, when the schema does not validate.
  • Security and privacy: PII classes present, retention, logging rules, who may see model output.
  • Compliance: residency, audit trail, human-review obligations.
  • Integration surface: which system consumes the output, in what format, with what idempotency guarantees.
The classic mis-elicitation

"The model should be accurate" is not a requirement — it cannot be tested, so it cannot be met. Convert it: "On the 200-invoice labelled set, field-level extraction accuracy ≥ 98% with zero silent failures; every uncertain field is flagged rather than guessed." That single rewrite dictates the eval, the schema (nullable fields plus a confidence or flag), and the human-review path.

Solution architecture inputs you do not control​

Requirements often arrive already constrained by an existing architecture. Read them for what they forbid:

Constraint in the architectureConsequence
Only outbound HTTPS from the service tierAn MCP server over stdio must run co-located; remote MCP needs a network path
No third-party network egress from productionSelf-hosted tool execution; no Anthropic-hosted sandbox
Existing identity provider and least-privilege modelTool calls carry the caller's authorization, verified server-side — never the model's assertion
Existing queue/worker topologyBatch fits naturally; a long-lived agent session may not
Existing observability stackLog usage, stop_reason, model ID, and request IDs in the format that stack ingests

Quiz​

  1. Question 1

    A stakeholder writes the requirement as "the extraction should be accurate". Which rewrite makes it implementable and testable?

  2. Question 2

    An interactive chat surface sends the same 80,000-token policy corpus with every request. Finance wants the per-request cost down, and the product team will not accept any change to what the model sees. Which mechanism applies?