Go deeper when you need to
A clearer introduction.
The detail is still here.
Every page on this site, grouped: the product story, the SQL MCP server and each supported engine, the comparisons and the Tableau guides, and the technical documentation packaged with the running version.
Understand the product
Capabilities, worked examples, trust, pricing and the practical questions before you evaluate it.
- Clear answers from your business data | datapipelines.co
- AI data pipeline — authored by an agent, run under review
- Text-to-SQL agent — from a question to a pipeline you rerun
- Federated query and data virtualization — cross-database joins
- dp-lake — query Parquet and Iceberg on S3 with SQL
- FAQ — datapipelines.co, the MCP pipeline server, in questions
- Roadmap — shipped, planned, and later | datapipelines.co
- AI agent database access, made safe — scoped keys, read-only, audited
- Turn a SQL query into an API — published endpoints | datapipelines.co
- A client reporting API per client — datapipelines for agencies
- Embedded analytics without a warehouse — the API ships today
- The AI SQL assistant with governance — one analyst, five databases
- How it works — pipelines, federated joins, MCP | datapipelines.co
- Demo data — NYC taxi, rideshare, US trade, weather | datapipelines.co
- Pricing — free and open source | datapipelines.co
- Semantic layer, learned by the agent — open source | datapipelines.co
- Use cases — operations, product and analytics teams | datapipelines.co
Connect your agent and databases
The SQL MCP server, each supported engine's driver and setup, client configuration and the tool list.
- SQL MCP server — one database MCP server, eight engines
- Postgres MCP server — governed, read-only, self-hosted
- SQL Server MCP server — governed, read-only, self-hosted
- MySQL MCP server — governed, read-only, self-hosted
- Oracle MCP server — governed, read-only, self-hosted
- SQLite MCP server — governed, read-only, self-hosted
- DuckDB MCP server — governed, read-only, self-hosted
- H2 MCP server — governed, read-only, self-hosted
- dp-lake MCP server — governed, read-only, self-hosted
- Add an MCP server to Claude Code, Cursor or Copilot
- The MCP tools — what an agent can do against your databases
Choose the right approach
Honest comparisons with the tools you may already run, and how the product works beside Tableau.
- Apache Airflow alternative for SQL pipelines — when to use each
- dbt alternative without a warehouse — when to use each
- Using datapipelines with Tableau — what ships today, what is planned
- Publish a governed dataset for Tableau from several databases
- Fivetran or Airbyte vs datapipelines — sync or query at the source
- Do you need a data warehouse? When Postgres alone is not enough
- Tableau Prep or pipelines as code — a flow is a file, this is code
- Embedded dashboards, extracts to S3, alerts — the roadmap
Read the technical documentation
Deployment, configuration, security, the pipeline and template contracts, the REST and MCP surfaces.
- Documentation
- Deployment & Packaging Specification
- Environments — naming them, and telling the product how careful to be
- Configuration Reference
- Auth & Security Specification
- Datasources Specification
- Key Providers — implementing a KMS-backed credential key source
- Observability Specification
- MCP Server Specification
- Pipeline Contract Specification
- Templates Specification
- Versioning: Draft, Release, Promotion
- REST API + SSE Specification
- Type System Specification
- Enumerations Reference
- Staging (H2) Specification
- DAG Executor Specification
- Calculators
- Metadata Database Schema Specification
- Module Structure Specification
- Pipeline Editor UI Specification
- UI Screens Inventory
- ROADMAP
- datapipelines.co — Specification Index