Open Source Mode Alternatives

Evidence is an open source, code-based alternative to drag-and-drop business intelligence tools. Reports are written as SQL queries and markdown, so a dashboard becomes a file in a repository rather than a layout somebody rebuilt by hand inside a vendor's web app.

The project describes itself as agent ready. You can develop inside the Evidence agent, or on your own machine alongside coding assistants such as Claude Code or Cursor, with a command line tool handling the rest of the loop.

The workflow stays close to the terminal.

• Install script: one command sets up the command line tool on macOS, Linux or Windows.
• Project scaffolding: evidence init creates a project and evidence dev serves it while you write.
• SQL and markdown: report sources are plain text, so they diff, review and merge like code.
• Static output: the build produces a site you can serve from your own infrastructure.
• Evidence Studio: publish to the hosted service when you would rather not run anything at all.

Self-hosting asks for little more than somewhere to serve static files, which makes this a light option next to a full BI server. Anyone comfortable writing SQL can produce reports without learning another vendor interface, and the maintainers run a public Slack channel for questions.

Lightdash asks analytics to behave like software. A context layer defines trusted metrics, joins, permissions, business logic and caching once, and everything downstream reads from it, whether that is a dashboard, an AI agent, a data app, an embedded view or the MCP server.

Metrics, charts and dashboards live as files. Changes are previewed from the command line, validated in CI and reviewed in pull requests, so analytics moves through the workflow engineers already trust. Definitions come from dbt projects or from standalone Lightdash YAML pointed at a warehouse.

That single layer is what the rest of the product is built on.

• Conversational analytics: agents answer from the context layer, respect permissions and return queries you can inspect.
• Data apps: build reports, workbooks, slide decks and forecasting tools from a prompt, with permissions and auth already wired in.
• Embedded analytics: an SDK supports row-level security, user attributes and customer-facing permissions.
• Agent skills: install the skills and MCP server so a coding agent can preview and validate before anything lands.
• Warehouse adapters: BigQuery, Snowflake, Redshift, Databricks, Postgres, Trino and ClickHouse among others.

The codebase is a TypeScript monorepo, React, Mantine and Vite on the frontend, Node, Express and PostgreSQL behind it. Self-hosting runs on Docker or Kubernetes with published Helm charts, enterprise features need a license key, and Lightdash Cloud exists for teams that would rather not operate any of it.

Deepnote is a drop-in replacement for Jupyter, built around a notebook format that survives version control. The messy JSON of an ipynb file gives way to a readable YAML file, one notebook per file, carrying its integrations and settings with it, so diffs stay focused enough to review.

Compatibility is kept deliberately. Deepnote uses its own kernel built on the Jupyter kernel, so existing notebooks keep working and you can move between the two. Notebooks are assembled from blocks rather than plain code cells, and reactive execution re-runs dependent blocks whenever their inputs or data change.

The repository holds the open pieces of that system.

• Blocks package: TypeScript types and utilities for code, SQL, text, markdown, input, visualization, image and separator blocks, including Python code generation.
• Convert CLI: bidirectional conversion between the Deepnote format and Jupyter, Quarto, percent Python and Marimo, as a command line tool or a programmatic API.
• Editor extensions: published extensions bring Deepnote notebooks into VS Code, Cursor and Windsurf.
• Snapshots: a snapshot file separates outputs from source for a cleaner history, with a content hash recording which code produced each output.
• Roundtrip fidelity: notebooks from Colab, SageMaker, Kaggle or Azure ML keep their platform-specific metadata when exported back.

One npx command converts an existing notebook for opening in your editor. The open source project targets local work, while a cloud service adds team collaboration and managed compute, free for students and educators. Planned work includes running the cloud interface locally, a local AI agent and bring-your-own model keys.