What is Right-Sizing?
Right-sizing is the practice of tailoring what a capability exposes to match what its consumers actually need. This applies at multiple levels: shaping API response payloads, curating AI context windows, extracting focused capabilities from monolith APIs, and aggregating microservices into business-oriented interfaces.
Right-sizing patterns in Naftiko
- Right-size AI context – Use output parameters and JSONPath mapping to expose only the fields an AI task needs, reducing noise and improving relevance.
- Right-size microservices – Aggregate multiple microservice endpoints under one capability with standardized contracts, hiding service fragmentation from consumers.
- Right-size a monolith – Extract focused capabilities from a broad monolith API so each consumer scenario gets a purpose-built interface.
- Right-size API content – Shape response payloads into smaller, typed outputs aligned to tasks rather than raw provider data dumps.
Why it matters
Over-exposure is the default in most integration architectures. Right-sizing reverses this by making precision the default – exposing the minimum needed for each consumer scenario. For AI consumers, this directly impacts accuracy and cost. For API consumers, it improves developer experience and reduces coupling.
