
Automated Insights
Transform raw data into human-scale narrative at enterprise speed.

The open-source framework for building high-quality data products with SQL and Markdown.

Evidence is a pioneering 'Analytics-as-Code' framework designed for data teams who prefer developer-centric workflows over drag-and-drop BI tools. Architecturally, it functions as a static site generator (SSG) that compiles SQL queries and Markdown files into high-performance, interactive data applications. By 2026, Evidence has positioned itself as the industry standard for version-controlled reporting, leveraging an internal DuckDB-Wasm engine to enable lightning-fast, client-side data exploration without repeated database round-trips. The platform bridges the gap between raw data and polished executive reports by treating dashboards like software projects—complete with Git integration, CI/CD pipelines, and modular component libraries. This approach eliminates the 'black box' nature of traditional BI, allowing for granular audit trails, automated testing of data visualizations, and seamless embedding into existing web ecosystems. Its 2026 market position is defined by its displacement of legacy tools like Looker for teams that prioritize speed, reproducibility, and the ability to maintain dashboards using standard IDEs.
Evidence is a pioneering 'Analytics-as-Code' framework designed for data teams who prefer developer-centric workflows over drag-and-drop BI tools.
Explore all tools that specialize in automated reporting. This domain focus ensures Evidence delivers optimized results for this specific requirement.
A cross-database SQL dialect that uses DuckDB to query multiple sources (databases, CSVs, Parquet) in a single statement.
Compiles data into a high-performance columnar database that runs inside the user's browser.
Uses file-based routing and Markdown templates to generate hundreds of unique pages from a single template (e.g., /customer/[id]).
Visualizations are built as reactive Svelte components, allowing for deep customization via CSS and JS.
A robust command-line interface for scaffolding projects, managing plugins, and automating builds.
Entire report structure, logic, and visualization code are stored as plain text files in Git.
Only re-runs queries and regenerates pages where the underlying data or code has changed.
Install Node.js (v18 or higher) on your local machine.
Run 'npx extra-evidence@latest create' to initialize a new project directory.
Navigate to the project folder and run 'npm install' to fetch dependencies.
Configure the 'sources' directory by adding a connection profile for your database (Postgres, Snowflake, BigQuery, etc.).
Define data environment variables (.env) for secure authentication credentials.
Create a new .md file in the 'pages' directory to start building a report.
Embed SQL blocks inside Markdown using the <sql> tag syntax to fetch data.
Use built-in Svelte-based components like <AreaChart /> or <DataTable /> to visualize the SQL results.
Run 'npm run dev' to launch the local development server with hot-reloading.
Execute 'npm run build' to generate the optimized static site for deployment to Evidence Cloud, Vercel, or Netlify.
All Set
Ready to go
Verified feedback from other users.
"Users highly praise the developer experience and speed, noting that it makes BI feel like modern web development. Some non-technical users find the SQL/Markdown requirement a steep learning curve."
Post questions, share tips, and help other users.

Transform raw data into human-scale narrative at enterprise speed.

Agentic Infrastructure for Personalized Experience

AI-powered solutions for enhanced resonance.

The Autonomous AI SOC Analyst for Security Operations

Enterprise Data Observability for Reliability, Cost Governance, and AI Pipeline Trust.

The world’s most comprehensive repository of computing research and citation data.