# Medallion architecture explained: Bronze, Silver and Gold

> Medallion architecture organizes lakehouse data into Bronze (raw, immutable copies of sources), Silver (cleaned, standardized and validated) and Gold (certified, business-ready tables). It makes pipelines auditable and reprocessable, and on DataNature it runs on open Delta or Iceberg tables with lineage in Atlas.

Source: https://datanature.ai/blog/medallion-architecture-explained
Publisher: Dlytica Inc. (DataNature is a Dlytica company)
Updated: 2026-11-05

## What belongs in Bronze?

Raw, append-only copies of each source, timestamped and partitioned. Never deleted, so you can always reprocess.

## What belongs in Silver?

Deduplicated, standardized data with nulls handled, referential integrity checked and business rules applied.

## What belongs in Gold?

Dimensional and fact tables, metrics and features that teams and AI agents consume. Certified and versioned.

## How do you promote data between layers?

With tested dbt or Spark jobs, run by Jobs and Pipelines, with every run writing lineage to [Atlas](/platform/atlas).

## How is the medallion architecture implemented in DataNature?

In DataNature each layer has a concrete home and a concrete tool:

| Layer | Where it lives | Built by | Quality rules |
|---|---|---|---|
| Bronze | `lakehouse.bronze_<source>.<table>` (Iceberg) | Connect (batch) and StreamHouse (real time) | Raw, immutable, all columns nullable, column names lowercased |
| Silver | `lakehouse.<schema>_silver` | dbt models on Spark 4 or Trino | De-duplicated (latest row per key), typed, `not_null` and `unique` tests on keys |
| Gold | `lakehouse.<schema>_gold`, plus serving tables in DN Warehouse | dbt marts per domain | Facts and dimensions, `relationships` tests, certified and version-controlled |

A typical job runs Connect pipeline → dbt Silver → dbt Gold → Refresh cubes, on a schedule with a time zone. Three habits keep it healthy: declare Bronze tables as dbt sources so freshness and lineage work, use incremental models for large facts, and build semantic-layer cubes on Gold only. Atlas records each run, so lineage traces any Gold metric back to the Bronze row it came from. For the full workflow, see [dbt on Spark and Trino with Iceberg](/blog/dbt-spark-trino-iceberg-medallion-pipelines).

## How does DataNature Atlas keep this secure and compliant?

Governance is part of the architecture, not a separate tool bolted on later. [DataNature Atlas](/platform/atlas) is the platform's governance module:

- **Knows your data**: a catalog with automated PII detection and classification across Bronze, Silver and Gold.
- **Controls access**: role-based grants plus attribute-based policies, row filters and column masks (null, redact, partial, hash, year only).
- **Governs AI**: Voir agents are scoped principals; masks, row filters and PII guardrails apply to every tool call and answer.
- **Proves it**: lineage from source to metric to decision, and an audit log of who or what accessed which data.
- **Stays local**: Atlas runs inside your deployment, so policies, lineage and logs never leave your jurisdiction.

These controls support GDPR, PIPEDA, Quebec Law 25, CCPA, HIPAA and sector rules such as OSFI E-23. Read [DataNature Atlas in depth](/blog/datanature-atlas-governance-rbac-abac-pii-masking) or see [security and compliance](/security).
