AI agents and copilots
Plug Claude and your enterprise copilots into your tables over MCP
Native MCP protocol
Data Access
Data Product APIs
Query every table over SQL, GraphQL, REST, MCP and search: one endpoint per protocol
Publish your tables as data products and plug in your BI tools, applications and AI agents: Claude and enterprise copilots connect natively over MCP. Catalog, quality, metrics and consumption, everything is managed from the console.
Specifications
- SQL
- PostgreSQL protocol, simple and prepared queries
- GraphQL
- Generated schema, GraphiQL playground
- REST / OpenAPI
- Generated endpoints, built-in Swagger
- MCP
- URL and token provided by the console
- Search
- Full-text, BM25 scoring
- Authentication
- Bearer token, service account rights
Use cases
Modern web applications
GraphQL integration for reactive interfaces
Typed GraphQL queries
Analytics and BI
PostgreSQL connection for your existing tools
Compatible with Tableau, Power BI, DBeaver, DataGrip
In action
The AI agent plugged into your data
A business team connects an AI agent to its tables over MCP
- Get the data product's MCP configuration from the console, with its URL and token
- Declare it in Claude, in an agentic IDE or in the enterprise copilot
- The agent lists the data products, catalogs, schemas and tables its token grants access to
- It then queries a table, choosing its columns, a filter and a maximum number of rows
An AI agent plugged into real data, confined to the scope its token allows
The developer who waits for no one
A frontend developer wants to display sales data in an application
- Open the data product in the console and select the GraphQL protocol
- Test the generated query in the built-in GraphiQL playground
- Copy the JavaScript or Python integration snippet shown next to the query
- Ship to production on the same endpoint, with a dedicated service account
The feature ships without going through an integration team
The BI analyst who keeps their tools
A BI analyst connects Tableau to Hyperfluid over the PostgreSQL protocol
- Get the PostgreSQL connection details from the console: host, port, database and token
- Enter them in Tableau, Power BI or DBeaver as a standard PostgreSQL source
- Build the dashboard from the tables exposed in the catalog
- Queries hit the tables directly, with no intermediate copy
A BI dashboard connected to the catalog, with no extra ETL pipeline
5 protocols, one source
- MCP
- Model Context Protocol: plug Claude, your AI agents and your enterprise copilots straight into your tables, with access rights specific to each token.
- PostgreSQL
- Connect your usual SQL tools over the PostgreSQL protocol: psql, DBeaver, DataGrip, Tableau, Power BI.
- REST / OpenAPI
- Endpoints generated from the schema, documented and explorable from the built-in Swagger UI.
- GraphQL
- Flexible queries for your modern frontend applications, with a GraphiQL playground in the console.
- Full-text Search
- Text search across your tables, BM25 scoring, index built on demand or pre-computed.
Governed, observable data products
Every data product is managed from the console, from publication to consumption tracking.
Catalog
All your tables published as data products, ready to consume.
Search
Find the right data product through catalog search.
Quality
Keep an eye on the reliability of each data product.
Metrics
Measure the activity of your data products.
Consumption
See who consumes your data, over which protocol and from which account.
Key benefits
- Five protocols, one endpoint per protocol, a single source
- Native MCP for your AI agents and enterprise copilots
- Quality, metrics and consumption tracked in the console
- Every access goes through a token, carrying the rights of the service account behind it
Ready to connect your applications?
Discover how Data Product APIs power your applications and AI agents.