Technologies

Native code for your platform, running in your cloud

VaultSpeed generates code in the dialect and objects of each target platform from one data model, and it deploys and runs like code your own team wrote.

Book a working session How the code is generated
One data model, many platforms

You run more than one platform

Most data teams run more than one platform: a cloud warehouse or lakehouse for new work, and SQL Server or Oracle on premises still doing important jobs. A platform strategy rarely ends on a single vendor. VaultSpeed generates for each target from the same data model, lineage and logic, and for two of them at once during a transition.

Native means the code is generated for the target, lands in your Git and CI/CD and runs without VaultSpeed in the loop, on open storage formats such as Iceberg and Delta where you use them.

Snowflake Select partner

Snowflake SQL, Tasks and Dynamic Tables from one data model

VaultSpeed is a Snowflake Select partner and Snowflake Ready validated technology. It runs as a native Snowflake application inside your trust boundary, with metadata only access, and is installed from the Snowflake Marketplace.

Native Snowflake artifacts

Snowflake SQL for DDL and DML, Tasks for scheduling and Dynamic Tables for incremental processing, with dbt models as an option.
Native Snowflake artifacts

Open storage, open metadata

Apache Iceberg tables where you want open storage, and metadata that keeps your Snowflake environment interoperable.
Open storage, open metadata

Controlled change

Incremental and CDC loading with deletes and reprocessing handled, and delta only deployments as your environment evolves.
Controlled change
Snowflake Select PartnerSnowflake Ready TechnologySnowflake Marketplace listing

Agents can use Snowflake Cortex through its REST API; model choice is a configuration, not an architecture decision.

The same context feeds Snowflake's own agents.

Context storeOne data model, lineage and logicWhat they answer traces back to the same graph.
MCP
Cortex AnalystAnswers from the semantic views generated from the context store
Cortex CodeWorks from the same data model, lineage and logic over MCP

How inference is set upHow the code is generatedWhere it runs

Databricks Validated Technology Partner

Unity Catalog objects and Spark SQL from one data model

VaultSpeed is a Databricks Validated Technology Partner. It generates native code for your lakehouse and integrates with Databricks Workflows, so schema drift is handled inside the platform you already run.

Unity Catalog and Spark SQL

DDL that exposes the data model in Unity Catalog, and transformation logic as Spark SQL notebooks or dbt models, across bronze, silver and gold.
Unity Catalog and Spark SQL

Delta Lake and Iceberg

Incremental and CDC loading on Delta Lake with deletes and reprocessing handled, and Apache Iceberg where you prefer it.
Delta Lake and Iceberg

Governance stays in Unity Catalog

Design time metadata lives in VaultSpeed; runtime governance stays centralized in Databricks and Unity Catalog.
Governance stays in Unity Catalog
Databricks Validated Technology Partner

Agents can use Databricks Model Serving, governed by Unity Catalog; model choice is a configuration, not an architecture decision.

The same context feeds Databricks' own agents.

Context storeOne data model, lineage and logicWhat they answer traces back to the same graph.
MCP
GenieAnswers from the semantic views the generator produces in Unity Catalog
Genie CodeWorks from the same data model, lineage and logic over MCP

How inference is set upHow the code is generatedWhere it runs

Generated dbt projects

A dbt project you would have written, minus the writing

Every model, test and description is generated from the context store, as plain dbt models or on the open source Data Vault packages your team already uses: datavault4dbt by Scalefree and AutomateDV. Every change is a pull request your team reviews.

Models and tests

Staging, raw vault, business vault and marts as dbt models, with uniqueness, relationship and freshness tests generated alongside.
Models and tests

Packages

Plain dbt models, or datavault4dbt by Scalefree and AutomateDV, formerly dbtvault, with the macros those packages expect.
Packages

What stays yours

Your macros, custom models and CI, untouched. An existing dbt project can be read into the context store and kept running.
What stays yours

See it live

Generated dbt Data Vaults, with the package maintainers

Two recorded sessions, one per open source package. Same context store, same review flow, different macros underneath.

Webinar · datavault4dbt

VaultSpeed and Scalefree

From source analysis to a reviewed pull request, generating a Data Vault on the datavault4dbt package with the Scalefree team.

Watch the recording

Webinar · AutomateDV

VaultSpeed and AutomateDV

The same generated project, this time on the AutomateDV macros, formerly dbtvault, with the people who maintain it.

Watch the recording

Customer story · Retail

Thanks to VaultSpeed, we industrialized our data layer without compromising agility or compliance. Our teams build and deploy data products faster, smarter, and at scale.
Colruyt GroupColruyt Group rebuilt and modernized its warehouse on Snowflake, at very high volumeRead the story

The rest of the landscape

Microsoft, Google, AWS, Oracle, SQL Server and SingleStore

Microsoft Fabric

Native code for Fabric, so teams on the Microsoft stack work from the same context as everyone else.

Google BigQuery

GoogleSQL generated to run directly in BigQuery, with lineage and definitions open to your catalogs.

Amazon Redshift

Redshift SQL generated for provisioned or serverless clusters, next to the rest of your AWS stack.

Oracle

PL/SQL and DDL generated for Oracle platforms, as the current platform or as the source of a migration.

Microsoft SQL Server

T-SQL generated for SQL Server, on premises or in Azure, from the same data model as the cloud targets.

SingleStore

SQL generated for SingleStore, so a distributed engine built for real time analytics works from the same data model as the rest of your landscape.

Leaving a legacy platform? See the use case

Talk to us

Start with the platform you run

Tell us what you run today and where it is heading, and we will show you the code VaultSpeed generates for it, with your integrator in the room if you want them there.