API & AI Agents
Automate beVault with APIs and AI agents.
beVault gives your teams a graphical interface for designing, documenting and managing Data Vault environments. But you are not limited to the interface. Use the beVault API to connect the platform to your applications, scripts and workflows. Use the beVault MCP Server to connect beVault to any AI agents.
Three ways to interact
One platform. Three ways to interact.
beVault adapts to the way your teams work.
Graphical interface
Use the beVault interface to model business concepts, map source systems, create Data Vault objects, configure data quality rules and manage Information Marts.
API integration
Use the API to automate operations, connect beVault to other enterprise tools and integrate the platform into your existing workflows.
MCP Server
Connect beVault to an AI agent and allow it to consult or create objects in your data model within the permissions and controls you define.
The interface, API and MCP Server are not separate products. They are complementary ways to interact with the same beVault platform.
One platform. Three ways to work: through the interface, through code and through AI.
API
The beVault API
Use beVault programmatically
The beVault API allows your applications and automation tools to interact with beVault without requiring users to perform every operation manually through the graphical interface.
The API provides programmatic access to the actions available in the beVault interface. This means that operations performed by a user can also be integrated into scripts, applications, workflows and enterprise platforms.
You can use the API to:
- Create and update entities in beVault.
- Synchronize metadata and documentation with your data catalog.
- Integrate beVault into CI/CD and deployment processes.
And much more: every action you can perform in the interface can be reproduced with the API.
Connect beVault to your existing ecosystem
beVault does not need to operate as an isolated platform. Use the API to connect it to the tools your organization already uses.
This allows beVault to become part of your broader data architecture rather than another disconnected application.
- Data catalogs and business glossaries.
- Data governance tools.
- Git repositories and CI/CD platforms.
- Workflow and orchestration tools.
- Internal applications.
And any other tool that can communicate through an API.
Automate repetitive modeling tasks
Many data teams repeatedly create similar objects and structures. With the beVault API, these operations can be automated.
For example, you can define a template for a recurring business concept and use it to create the corresponding entities in beVault automatically. This helps standardize implementation, reduce manual effort and accelerate the delivery of new business cases.
A typical automation could:
- 1Read metadata from a source system.
- 2Apply your modeling rules or template.
- 3Create the required entities in beVault.
- 4Update the data catalog and technical documentation.
- 5Trigger the next step in your delivery workflow.
Automation
From manual repetition to reusable automation
Instead of rebuilding the same structures by hand, your team can create reusable processes that produce consistent results every time.
MCP Server
Connect beVault to AI agents with MCP
The Model Context Protocol, or MCP, provides a standard way for AI applications to connect to external tools and data sources.
The beVault MCP Server allows MCP-compatible AI agents to interact with your data model within the permissions and controls you define.
For example, an AI agent could answer:
“Which business concepts are connected to the customer domain?”
Or:
“Create the standard satellite structure for this source using our template.”
Depending on their access, agents can:
- Answer questions about your data model.
- Explore metadata, entities, relationships and lineage.
- Create or update objects in beVault.
- Assist with repetitive modeling tasks.
- Help you build Information Marts from your existing model.
And any other task you give your agent, within the access you grant it.
Interactive demo: starting from an empty beVault project, the agent proposes a data model for the user to review and validate before it is executed.
Next step