03-input-data-building-model.md

03 - Input Data Building Model

This is the most important model shape to understand, because it captures the core mental model of the whole ecosystem.

Raw Input Data
  ↓
Normalize
  ↓
Data State 1
  ↓
Extract / Pick
  ↓
Data State 2
  ↓
Structure
  ↓
Built Data
  ↓
Evaluate
  ↓
Feedback

The key distinction

• Data is not the transform.
• Data flows through transforms.
• The graph turns raw state into useful state.
• The trace exposes each intermediate state.

A transform is a verb (normalize, extract, structure). Data is a noun (the raw text, the clean text, the extracted words, the built record). The graph is the ordered composition of the verbs; the trace is the recording of every noun along the way.

Formula

D0 = raw input data
D1 = normalize(D0)
D2 = extract(D1)
D3 = structure(D2)

BuiltData  = D3
Evaluation = E(D3)

In one line

Input Data Building is the process of turning raw state into useful state through explicit, inspectable transforms.

Worked example

Input:

{ "text": "  Composable Model Graph builds explicit transformations  " }

Transforms:

• normalizeText — trim, lowercase, collapse whitespace
• extractWords — split into words
• buildWordData — structure into { words, count }

Output:

{
  "words": [
    "composable",
    "model",
    "graph",
    "builds",
    "explicit",
    "transformations"
  ],
  "count": 6
}

Evaluator: pass if count > 0.

Trace:

RawInput      { text: "  Composable Model Graph builds explicit transformations  " }
   ↓ normalizeText
CleanData     "composable model graph builds explicit transformations"
   ↓ extractWords
ExtractedData ["composable","model","graph","builds","explicit","transformations"]
   ↓ buildWordData
BuiltData     { words: [...], count: 6 }

See the runnable version in example 05.