Data-flow edges specify values, provenance, and schemas exchanged between nodes. Separating them from control flow prevents a value's presence from silently authorizing execution.

Data Flow Answers What Moves

Chunking divides sources into retrievable units. Good boundaries preserve claims, definitions, qualifiers, and provenance while avoiding fragments too broad or too small to answer reliably.

Chunking Preserves Answerable Context

In this situation—passing retrieved passages and citations into an answer node—which choice best applies “Data Flow Answers What Moves”?

A map stage applies comparable work to independent partitions; a reduce stage combines results under explicit rules. Partition boundaries and reconciliation determine whether context is lost.

Map-Reduce Splits and Reconciles Work

Prompt chaining decomposes work into model nodes whose outputs become validated inputs to later nodes. Each boundary should add a distinct transformation, check, or decision.

Learn Before

Related