Skip to content

AIP-C01 · Study Guide

← Domain 1: Foundation Model Integration, Data Management, and Compliance

1.3.3 · Lesson 10 of 28

Format input data for FM inference

Transform data into the exact input structure expected by the selected model/API.

WHAT

Transform data into the exact input structure expected by the selected model/API.

Examples

  • JSON request body
  • structured fields
  • conversation messages
  • multimodal content blocks
  • endpoint-specific request format

WHY

Correct data with incorrect formatting can still produce failed inference.

AWS

Amazon Bedrock APIs generally require model/API-specific request structures.

SageMaker endpoints can require model-specific inference payloads.

Best practice

Create an adapter:

Application Data
      ↓
Normalization
      ↓
Model Adapter
      ↓
Model-specific Request

EXAM TRAP

Do not assume one universal request format works for all models.

---

Key points

  • Correct data with incorrect formatting can still produce failed inference.

Exam focus

JSON request body structured fields conversation messages multimodal content blocks endpoint-specific request format

View progress