Lightdash: Open Source Data Engineering & Integration

What is Lightdash?

What it is

Lightdash is an open-source, agentic business intelligence platform for modern data teams, written in TypeScript and distributed under a NOASSERTION license. It lives in the data and analytics ecosystem, specifically the data engineering and integration space, and it is built around dbt and warehouse-native workflows. The project describes itself as agentic BI, meaning analytics is intended to move at the speed of code rather than through point-and-click configuration alone.

The concrete problem it solves is governance drift between how metrics are defined and how they are consumed. Lightdash introduces a context layer where metrics, dimensions, joins, descriptions, caching, and access rules are defined once, then reused across dashboards, AI agents, data apps, embedded analytics, SDKs, and MCP. Because those definitions live as files, they can be built with coding agents, previewed from the CLI, validated in CI, and reviewed in pull requests. Business users can then ask questions in plain English or build data apps without bypassing the governance that the context layer enforces.

Key capabilities

BI as code: metrics, charts, and dashboards live as files that can be previewed from the CLI, validated in CI, and reviewed in pull requests.
AI agents that answer from the context layer rather than raw table guesses, respecting permissions and returning inspectable queries.
Data Apps for building custom reports, workbooks, slide decks, forecasting tools, and customer-facing data products from a prompt.
A context layer defining metrics, dimensions, joins, descriptions, caching, and access rules, sourced from dbt projects or standalone Lightdash YAML pointed at a warehouse.
Embedded analytics through the Lightdash SDK, with configurable row-level security, user attributes, and customer-facing permissions.
Agent skills and an MCP server for building charts, dashboards, metrics, and Data Apps from an editor or terminal.
An open-source core that can be self-hosted, with commercial enterprise features available under a license key.

Who uses it and how

Data teams that maintain analytics as code, using the terminal, pull requests, and CI to preview, validate, and ship changes.
Teams connecting dbt projects or standalone Lightdash YAML to a warehouse to define a single governed context layer.
Business users asking questions in plain English, exploring dashboards, or creating custom data apps within governance boundaries.
Product teams embedding dashboards, AI agents, and Data Apps into their own products with row-level security and user attributes.
Organizations running Lightdash on their own infrastructure with Docker or Kubernetes, optionally adding enterprise features via a license key.

Getting started

The fastest path is Lightdash Cloud, a hosted workspace with no infrastructure to run; a live demo and documentation are also available. Self-hosting runs Lightdash on your own infrastructure with Docker or Kubernetes, and production deployments should begin with the production deployment checklist.

When to use it — and when not to

Lightdash fits teams already invested in dbt or warehouse-native modeling that want analytics governed through Git, CI, and pull requests rather than configured only in a UI. A self-hoster must operate the deployment itself, whether on Docker or Kubernetes, and should work through the production deployment checklist before going live; enterprise features on self-hosted deployments require a license key obtained from the Lightdash team. The repository carries a NOASSERTION license and a large open issue count, so teams should review licensing terms and issue backlog before committing to the open-source core.

project readme (upstream, from github) — read inline

Open-source Agentic BI for modern data teams

Build governed metrics, dashboards, AI agents, and data apps from your context layer. Ship analytics through Git, CI, MCP, and the Lightdash CLI.

Why Lightdash

Lightdash is the Agentic BI platform for teams that want analytics to move like software. Your context layer defines trusted metrics, joins, permissions, business logic, and caching once, then powers every way people consume data: dashboards, AI agents, data apps, embedded analytics, SDKs, and MCP.

Data teams can build and maintain analytics with any AI agent, then use the terminal, pull requests, and CI to preview, validate, and ship changes. Business users can ask questions in plain English, explore dashboards, or create custom data apps without bypassing governance.

Start with Cloud

The fastest way to use Lightdash is Lightdash Cloud: no infrastructure to run, always up to date, and ready for AI agents, Data Apps, scheduled deliveries, embedding, and enterprise controls.

Want to explore first? Try the live demo, read the docs, or book a sales call.

What you can build

BI as code

Your metrics, charts, and dashboards live as files. Build them with coding agents, preview changes from the CLI, validate in CI, and review analytics in pull requests.

Learn about BI as code

AI agents

Lightdash agents answer from your context layer, not raw table guesses. They respect permissions, return inspectable queries, reuse verified answers, and improve through reviews and evaluations.

Explore Conversational Analytics

Data Apps

Build custom reports, workbooks, slide decks, forecasting tools, and customer-facing data products from a prompt, with your context layer, permissions, and auth already built in.

See Data Apps

Context layer

Define metrics, dimensions, joins, descriptions, caching, and access rules in one governed layer. Use dbt projects or standalone Lightdash YAML pointed at your warehouse.

View the context-layer visual

Embedded analytics

Embed dashboards, AI agents, and Data Apps in your product with the Lightdash SDK which supports configurable row-level security, user attributes, and customer-facing permissions.

Take a look at our embedding showcase or read our embedding docs.

Open-source core

Self-host the core BI platform, contribute improvements, and run Lightdash in your own infrastructure. Enterprise deployments can add commercial features and support.

Self-host Lightdash

Build with agents

Lightdash gives coding agents the context they need to make safe analytics changes: install the Lightdash skills, preview what changed, then validate the project before anything lands.

lightdash install-skills
lightdash preview
lightdash validate

Use agent skills and the Lightdash MCP server to build charts, dashboards, metrics, and Data Apps from your editor or terminal. Every change can still go through the workflow your team trusts: preview the branch, validate the project, review the pull request, merge.

Installation

Cloud

Sign up for Lightdash Cloud to get a hosted workspace in minutes. This is the recommended path for teams that want AI agents, Data Apps, managed upgrades, and production-ready infrastructure without operating Lightdash themselves.

Self-host

Run Lightdash on your own infrastructure with Docker or Kubernetes. If you're deploying to production, start with the production deployment checklist. Enterprise features are available on self-hosted deployments with a license key; speak to the Lightdash team to get one.

Local development

git clone https://github.com/lightdash/lightdash.git
cd lightdash
./scripts/install.sh

See the contributing guide for the full local setup, package scripts, and development workflow.

Stack

Lightdash is a TypeScript monorepo built with:

React, Mantine, Vite, and TanStack Query on the frontend
Node.js, Express, TSOA, Knex, and PostgreSQL on the backend
Warehouse adapters for BigQuery, Snowflake, Redshift, Databricks, Postgres, Trino, ClickHouse, and more
A CLI, MCP server, SDKs, Git integrations, and content-as-code workflows for developer-native analytics

Community

Lightdash is open source and shaped by community feedback.

Thanks to everyone who has contributed code, docs, issues, ideas, and feedback. See the contributors graph for the full community.

Frequently asked questions

Is Lightdash free to use?

Lightdash is open source. There is no licence fee and no seat count — you can self-host it or, where the project offers one, pay a vendor for a managed version instead.

What does Lightdash do?

Open-source BI for modern data teams

What is Lightdash written in?

Lightdash is primarily written in TypeScript. Its source is publicly available at https://github.com/lightdash/lightdash, and it has 6,139 GitHub stars.