BoltPipeline logo

Enterprise Model

Your organization-wide data catalog — automatically built and kept current as you certify pipelines. No separate catalog tool required.

What it is

A catalog that builds itself

The Enterprise Model is BoltPipeline's organization-wide view of your data estate. It catalogs every table your agents discover — with SCD classification, layer assignment, business description, column-level metadata, producer/consumer relationships, and coverage status.

Unlike standalone catalog tools that require manual annotation and constant maintenance, BoltPipeline's Enterprise Model is derived automatically from two sources: the schema and statistics your agents discover in your warehouse, and the pipeline certifications you approve. As you certify pipelines, the catalog updates itself — no maintenance burden.

What the Enterprise Model captures per table

Table layer (Bronze, Silver, Gold, or custom)
SCD type (0 = fixed, 1 = overwrite, 2 = history)
Materialization (table, view, incremental)
Business description and grain
Owner and responsible team
Tags and domain classification
Column names, types, roles, and semantic type
PII classification per column
Data quality scores and null rates
Producing pipeline and last certification
All consuming pipelines
Sync status and last profiled timestamp

The problem it solves

Tribal knowledge is a governance liability

In most data teams, critical knowledge about the data warehouse lives in people's heads — which table is SCD Type 2, which pipeline produces the customer dimension, whether that table in the reporting schema is still actively maintained. When those people leave, the knowledge leaves with them. BoltPipeline eliminates this problem by making governance metadata machine-readable and always current.

Common tribal knowledge questions

  • ?Which pipeline is writing to orders_summary?
  • ?Is dim_customer SCD Type 2? What columns are the history keys?
  • ?If I change this source table schema, what breaks?
  • ?Which tables in our warehouse have no active pipeline maintaining them?
  • ?What raw data do we have that we haven't built analytics on yet?

With the Enterprise Model

  • The producer pipeline is shown on every table's detail page
  • SCD type and governance policy are stored, versioned, and auditable
  • Impact analysis shows every downstream consumer instantly
  • Orphaned tables (no active producer) are flagged automatically
  • Coverage map shows what's in the warehouse vs. what's governed

Governance layer

SCD type is a governance contract, not just metadata

When you classify a table as SCD Type 2 in the Enterprise Model, that classification becomes a governance constraint that BoltPipeline enforces across the entire platform. It is not a sticky note — it actively shapes what pipelines can and cannot do with that table.

SCD conformance is validated at certification

The pipeline that writes to an SCD Type 2 table must produce rows that preserve history — inserting new rows with effective date ranges instead of overwriting. BoltPipeline validates this at certification. A pipeline that produces overwrite (Type 1) semantics against a Type 2 table does not pass.

Consumer pipelines see the full contract

When a pipeline reads from a Type 2 table, it inherits the knowledge that the table contains multiple rows per key (one per historical version). BoltPipeline surfaces this during planning so developers write their SQL correctly — filtering to the active row, or intentionally querying historical state.

Table ownership is visible before you build

BoltPipeline tracks which certified pipeline writes to each table. Before anyone starts building a pipeline that targets that table, the Enterprise Model shows the existing producer — so teams can coordinate rather than discover conflicts after a production failure.

Change impact is immediate and precise

If an admin changes a table's SCD classification in the Enterprise Model, BoltPipeline immediately surfaces every pipeline that produces or consumes that table. You see the full impact before certifying the change — not after a Production failure.

Lineage

Every table knows who builds it and who uses it

The producer/consumer graph is built automatically as you certify pipelines. No manual annotation. No catalog maintenance. Every certification records which tables a pipeline reads from (consumers) and which tables it writes to (producer). BoltPipeline aggregates this into a complete map of your data estate.

Example: producer/consumer map for dim_customer

Producer

CRM Pipeline

Certified · SCD Type 2

Table · SCD Type 2

analytics.dim_customer

Silver Layer · 2.4M rows · Healthy

Consumer

Customer 360 Pipeline

Consumer

Finance Reporting Pipeline

Consumer

Churn Model Pipeline

BoltPipeline builds this graph automatically. No manual wiring. Certifying a new pipeline that reads from dim_customer instantly adds it as a consumer.

Producer tracked per table

BoltPipeline records which certified pipeline writes to each managed table. Any team building a new pipeline that targets that table can see the existing producer — eliminating silent competition and making ownership unambiguous.

Many consumers tracked

Any number of pipelines can read from a table. All are tracked. When the table or its producer changes, BoltPipeline shows every downstream consumer that may be affected.

Orphan detection

Tables with no active producer are flagged as orphaned. Tables with no active consumers are flagged as unused. Both are visible in the Enterprise Model coverage view.

Real-time impact analysis

Before certifying any change that touches a table, BoltPipeline shows the full upstream/downstream blast radius — so you understand the consequences before they reach production.

Data inventory

See what you have — and what you're missing

Your data warehouse is an inventory of analytical raw material. Most organizations know roughly what they have — but cannot tell you with confidence which of those tables are being actively transformed into business insights, which have been abandoned, and where the unexploited opportunities are.

The Enterprise Model gives you this map. You can see every table your agent has discovered, classified alongside its governance status — whether it has an active certified producer, how many pipelines consume it, when it was last profiled, and whether it contains PII that requires handling.

The coverage question every data leader should be asking

What percentage of your warehouse tables have a certified pipeline maintaining them?

Coverage score

How many tables exist in your warehouse that no current pipeline reads from or writes to?

Orphan count

Where is raw source data sitting untransformed — the analytics you haven't built yet?

Opportunity map

Which tables contain PII that your current pipelines are not handling correctly?

Compliance exposure

Architecture

Layer your data like an enterprise

The Enterprise Model supports standard data lakehouse layering — Bronze, Silver, and Gold — as well as custom layer definitions. Every table is assigned to a layer, giving you an architectural map of your data estate alongside the governance metadata.

Bronze

Raw data exactly as it arrives from source systems. No transformation, no business rules. The full historical record — append-only or SCD Type 0.

Silver

Cleaned, conformed, and lightly transformed. Standard dimensions and facts. SCD Type 1 or Type 2 based on business requirements. The layer most pipelines read from.

Gold

Business-ready aggregations, KPIs, and reporting models. Read-heavy. BI tools connect here. SCD Type 1 most common — consumers need current state.

One platform

No separate catalog tool needed

Data catalog tools like Alation, Collibra, and Atlan solve a real problem — but they solve it by sitting on top of your existing pipeline and warehouse infrastructure. You buy them separately, integrate them manually, and maintain them continuously. And because they are disconnected from the pipeline lifecycle, their metadata is always one step behind.

CapabilityStandalone catalogBoltPipeline
Table & column metadataManual annotationAuto-populated by agents
SCD governance enforcementNot availableBuilt in to certification
Producer/consumer mappingManual / partialAutomatic from pipeline certifications
Stays currentRequires manual syncUpdates on every certification
Integrated with pipeline opsSeparate productSame platform
Coverage / gap analysisLimitedFirst-class feature
PII detectionAdd-on or separate toolBuilt in to agent profiling

Turn SQL into Production-Ready Data Pipelines — Faster and Safer

SQL-first pipelines, validated and governed — executed directly inside your database.

No new DSLs. No fragile orchestration. Just SQL with built-in validation, lineage, and governance.