dbt_salesforce_analytics_vault

A dbt package that builds a Data Vault 2.0 model of your Salesforce data. It takes the raw Salesforce tables delivered by the Fivetran Salesforce connector, turns them into a raw vault (hubs, links, satellites) using Scalefree's datavault4dbt macros, and then exposes analytics-ready delivery models on top.

Supported platforms:

Currently supported platforms are all databases that the Fivetran Salesforce Connector can write to and datavault4dbt supports.

Layers

The models are organized into three layers:

Folder	Layer	What it does
models/01_stage	Staging	Wraps the Fivetranstg_salesforce__* models and computes hash keys / hashdiffs.
models/02_raw_vault	Raw Vault	The Data Vault: hubs, links and satellites.
models/03_information_delivery	Information Delivery	Analytics-ready views on top of the data vault (e.g.salesforce__opportunity_enhanced, salesforce__owner_performance, salesforce__daily_activity, salesforce__sales_snapshot).

Entities modeled

Account, Contact, Event, Lead, Opportunity, Opportunity Line Item, Order, Product, Task, User and User Role.

The diagram below shows the resulting raw vault — hubs, links and satellites per entity. It was drawn with Visual Data Vault, Scalefree's browser-based editor for Data Vault diagrams. Click the image for the full-resolution version.

Naming conventions

• _h = hub, _l = link, _hl = hierarchical link.
• Satellites come in two flavours:
  • v0 = raw, fully historized satellite; v1 = current-record view.
  • n = business (non-PII) attributes; p = personal (PII) attributes.

Requirements

• dbt (tested with dbt-core 1.11 / dbt-bigquery 1.11).
• Salesforce data landed by the Fivetran Salesforce connector.

Dependencies

Declared in packages.yml:

Package	Role
dbt-labs/dbt_utils	Utility macros / tests.
ScalefreeCOM/datavault4dbt	Data Vault macros.
fivetran/salesforce	Upstream Salesforce staging models.

Getting started

1. Install dependencies:

   dbt deps

2. Configure a profile named dbt_salesforce_analytics_vault in your profiles.yml.
3. Point the fivetran/salesforce package at the dataset(s) where Fivetran lands your Salesforce tables (see that package's documentation for its variables).
4. Build the models:

   dbt build

Opening this model in turbovault-engine

This package ships its Data Vault metadata as a turbovault-engine JSON export: docs/dbt_salesforce_analytics_vault_export.json. turbovault-engine is Scalefree's CLI-first generator that turns source metadata into a production-ready datavault4dbt project; it accepts its own JSON exports as input, so the model can be re-imported, inspected, extended and regenerated:

1. Install the engine:

   pip install turbovault-engine

2. Initialize a workspace (once per directory):

   turbovault workspace init

3. Import this package's export as a project (run from the package root so the relative path resolves):

   turbovault project init --name salesforce_dv \
     --source docs/dbt_salesforce_analytics_vault_export.json

4. Inspect and edit sources, stages, hubs, links and satellites in the web UI:

   turbovault serve

5. Regenerate dbt models after changes:

   turbovault generate --project salesforce_dv --output ./generated

See the turbovault-engine README for all options, and Scalefree's TurboVault page for background.