Skip to content

Architecture

The four-layer architecture

A trusted data platform needs four layers, not two. Most organizations are missing the one that matters most at decision time.

Layer 4CONSUMPTION

Humans + Agents

Dashboards, notebooks, SQL, AI agents, automated flows

The trust gap closes here or it does not close at all

Layer 3TRUTHTHE MISSING LAYER

Governs the confidence

Trust scores, citations, decision thresholds, freshness, ownership accountability

Layer 2SEMANTIC LAYER

Governs the meaning

Authoritative definitions, business logic, dimension hierarchies. "One metric, one meaning."

Layer 1DATA CONTRACTS

Governs the pipe

Schema validation, freshness SLAs, pipeline quality gates

Most platforms stop here(Layers 1 and 2)
The truth layer spans both(Layers 2 and 3)

The maturity of a metric

From raw signal to a number an agent can trust.

Most platforms resolve data into a defined metric and stop — three stages in. A number you can act on needs three more: lineage, a score, and calibration against outcomes. Maturity isn't more data. It's more structure, and more connection.

Phase one · raw → defined

01 · Ingest

Raw signal

mine · gather · measure

atoms

02 · Model

Structured data

join · shape · classify

molecules

03 · Define

Defined metric

define · certify · version

DNA

Phase two · the truth layer

04 · Link

Lineage-linked

trace · connect · sequence

chromosomes

05 · Score

Scored metric

score · weight · grade

cells

06 · Trust

Trusted layer

calibrate · act · decide

organisms

Maturity visualization after the DIKW model (Ackoff / Zeleny); “data → knowledge” treatment inspired by David McCandless.

Data contracts solved the plumbing. Schema validation, freshness SLAs, pipeline guarantees. They answer one question. Did the data arrive correctly?

Semantic layers solved the definitions. One metric, one meaning. Airbnb's Minerva, dbt's metrics layer, and the broader push toward governed metric definitions all address the same problem. They answer a different question. What does this number mean?

Neither layer answers the question that actually matters at decision time. Should I act on this number, and with what confidence? That is the truth layer. It sits on top of both and adds what has been missing. Validation status, confidence scoring, lineage that a business user can actually read, and freshness signals that tell you whether the number reflects reality right now or reality from three days ago.

Data contracts govern the pipe. Semantic layers govern the meaning. The truth layer governs the confidence.

Justin Nixon · The truth layer · 2026

Part of: Stage 01 · The Missing Layer

Back to the map

Data contracts govern the pipe. Semantic layers govern the meaning. Neither answers the question that decides everything at the moment of action: should I act on this number, and with what confidence? That is the truth layer, and almost no platform has built it.

Read the essay