Dataherald

Paid APISelf-hostedWeb

What Dataherald is

Dataherald is an open-source natural language-to-SQL engine that lets people ask plain-English questions about relational data through an API. It is aimed at developers building natural-language interfaces over structured data or contributing to the project. Listed use cases include helping business users query a data warehouse, adding data Q&A to SaaS products, and creating a ChatGPT plug-in for proprietary data. The project includes an engine, an enterprise API layer with authentication and organization and user support, an admin console for configuration and observability, and a Slackbot. Its quick start describes running the engine locally with Docker and Docker Compose, with MongoDB in a separate container for application data. Supported warehouse connections are PostgreSQL, BigQuery, Databricks, and Snowflake. Included SQL-generation approaches include LangChain SQL Agent and SQL Chain, LlamaIndex SQL Generator, and Dataherald SQL Agent. A context store can use scanned table and column details, descriptions, and verified question-and-SQL examples to guide generation. The documentation also describes active learning from usage and an evaluator that assigns generated SQL a confidence score. Startup requires an OpenAI API key, organization ID, and encryption key. Database connection data is encrypted before storage in Mongo using the configured key.

Who it suits

Dataherald suits developers building natural-language question interfaces over structured data, including API based data Q&A features. It may also suit developers who want to self-host or contribute to an Apache-2.0 licensed project.

What it does well

Exposes an API for plain-English questions about relational data.
Connects to PostgreSQL, BigQuery, Databricks, and Snowflake.
Includes four documented SQL-generation implementations.
Context can include table details and verified question-and-SQL examples.
Licensed under Apache-2.0, with contributions invited.

What to know first

Self-hosting requires Docker and MongoDB.
Startup requires an OpenAI API key, organization ID, and encryption key.
Generated SQL may be inaccurate until business logic and data context are added.

Getting started

Open the Dataherald GitHub repository.
Prepare Docker and Docker Compose, with MongoDB running separately for application data.
Set an OpenAI API key, organization ID, and encryption key in the environment configuration.
Start the engine and open the Swagger UI at /docs.
Connect a supported warehouse and add table and column details or verified question-and-SQL examples to the context store.

Verdict

Pick Dataherald if you are a developer building a natural-language interface over structured data and want an open-source engine you can self-host and configure. Its context store and multiple SQL-generation implementations support adaptation to a database and its business logic. Expect to supply the required credentials and build up context, since generated SQL may be inaccurate until that context is added.

Questions

Does Dataherald cost anything?

The project is identified as free and open source under the Apache-2.0 license.

Which databases can it connect to?

The quick start lists PostgreSQL, BigQuery, Databricks, and Snowflake.

Can I host Dataherald myself?

Yes. The quick start describes running the engine locally with Docker and Docker Compose, alongside MongoDB for application data.

What do I need before starting the engine?

The environment configuration requires an OpenAI API key, an organization ID, and an encryption key.

How does Dataherald use database context?

The context store can use scanned tables and columns, descriptions, and verified question-and-SQL examples to guide SQL generation.

Can generated SQL be inaccurate?

Yes. The quick start says generated SQL may be inaccurate until business logic and data context are added to the context store.

Facts from Dataherald's own pages

Adaptability: The documentation describes a modular architecture whose implementations can be replaced through configuration in the .env file. · dataherald.readthedocs.io, 4 Oct 2026
API: The engine exposes endpoints for database connections and natural-language questions, and the quick start points to a Swagger UI at /docs. · dataherald.readthedocs.io, 8 Oct 2026
Components: The repository describes an engine, an enterprise API layer with authentication and organizations, an admin console for configuration and observability, and a Slackbot. · github.com, 4 Oct 2026
Context: The context store can use business and database context, including validated question-and-SQL examples and table information, to help generate SQL. · dataherald.readthedocs.io, 4 Oct 2026
Context and learning: Users can add database context by scanning tables and columns, providing verified SQL examples, and describing tables and columns; the documentation also describes active learning from usage. · dataherald.readthedocs.io, 8 Oct 2026
Credentials: The quick start requires an OpenAI API key, an organization ID, and an encryption key before starting the engine. · dataherald.readthedocs.io, 4 Oct 2026
Database integrations: The quick start says the engine supports PostgreSQL, BigQuery, Databricks, and Snowflake connections. · dataherald.readthedocs.io, 8 Oct 2026
Database support: The quick start lists PostgreSQL, BigQuery, Databricks, and Snowflake as currently supported warehouse connections. · dataherald.readthedocs.io, 4 Oct 2026
Deployment: The quick start describes self-hosting the engine with Docker, with Mongo running in a separate container for application data. · dataherald.readthedocs.io, 4 Oct 2026
Deployment components: The repository includes an engine, an enterprise API layer with authentication and organization and user support, an admin console for configuration and observability, and a Slack bot. · github.com, 8 Oct 2026
Evaluation: The module overview says an evaluator assigns a confidence score to generated SQL. · dataherald.readthedocs.io, 4 Oct 2026
Intended users: The documentation is intended for developers building natural-language interfaces from structured data in their own projects and contributors to the engine. · dataherald.readthedocs.io, 8 Oct 2026
License: The GitHub repository identifies the project as licensed under Apache-2.0. · github.com, 8 Oct 2026
Notable accuracy consideration: The quick start says generated SQL may be inaccurate until business logic and data context are added to the context store. · dataherald.readthedocs.io, 8 Oct 2026
Product: Dataherald is a natural language-to-SQL engine that exposes an API for asking questions about relational data in plain English. · github.com, 4 Oct 2026
Requirements: The engine requires an OpenAI API key, an organization ID, and an encryption key in its environment configuration before startup. · dataherald.readthedocs.io, 8 Oct 2026
Security: The quick start says database connection data is encrypted before storage in MongoDB using an ENCRYPT_KEY and recommends changing the default MongoDB username and password. · dataherald.readthedocs.io, 8 Oct 2026
Self-hosting: The quick start describes running the engine locally with Docker and Docker Compose, alongside MongoDB for application data. · dataherald.readthedocs.io, 8 Oct 2026
SQL generation: Included text-to-SQL implementations are LangChain SQL Agent, LangChain SQL Chain, LlamaIndex SQL Generator, and Dataherald SQL Agent. · dataherald.readthedocs.io, 4 Oct 2026
SQL generation options: The documentation lists LangChain SQL Agent, LangChain SQL Chain, LlamaIndex SQL Generator, and Dataherald’s in-house SQL agent as included implementations. · dataherald.readthedocs.io, 8 Oct 2026
Support and licensing: The repository is public and identifies its license as Apache-2.0; the README invites contributions to the project. · github.com, 4 Oct 2026
Use cases: The project describes use cases including helping business users explore a data warehouse, adding Q&A to SaaS applications, and creating a ChatGPT plug-in from proprietary data. · github.com, 8 Oct 2026
What it does: Dataherald is a natural language-to-SQL engine that lets users ask questions in plain English about structured or relational data through an API. · github.com, 8 Oct 2026

Where Dataherald stands on Freedom251

AI Database Assistants#18 of 21
AI SQL Generators#17 of 19