Query, explore, and join across every database — Postgres to Snowflake to MongoDB — without switching tools or spinning up infra.
Relational, warehouse, NoSQL, and streaming — query them all side by side, and join across them in a single window.
A built-in DataFusion engine lets you write SQL across Postgres, Snowflake, MongoDB and more — joined in memory, no ETL pipeline needed.
SELECT
u.country,
SUM(o.amount) AS revenue,
COUNT(e.event_id) AS events,
MAX(s.last_seen) AS last_active,
AVG(f.predicted) AS forecast
FROM postgres.public.users u
JOIN snowflake.sales.orders o ON o.user_id = u.id
JOIN mongo.analytics.events e ON e.user_id = u.id
JOIN redis.cache.sessions s ON s.user_id = u.id
JOIN duckdb.main.forecasts f ON f.country = u.country
GROUP BY u.country
ORDER BY revenue DESC
Open your project. We parse your manifest, render lineage, and let you run, test, and ship models with the same shortcuts you'd expect from your editor.
Run a Python console or open a Jupyter notebook right next to your SQL.
Run a query, hit one keystroke, and get a chart.
Arris indexes your schema so the AI writes queries that actually run. Review the SQL, hit run, and move on.
Beyond the headline features, Arris ships the everyday essentials a data IDE should have.
Arris is a Mac desktop app, with native builds for both Apple Silicon and Intel. It is built on Tauri (a Rust core with a web-based UI), so it installs and runs like any other Mac app, with no website or browser tab to keep open.
Everything stays on your Mac. Connections, query results, and federated joins are processed locally with no cloud sync, so your data never leaves your machine.
Arris ships 18 bundled drivers spanning SQL, NoSQL, and streaming sources — Postgres, MySQL, Snowflake, MongoDB, Redis, DuckDB, SQLite, and more — each with a schema browser tuned to that engine.
An embedded DataFusion engine lets you write a single SQL statement that joins across multiple connections. Filters and projections are pushed down to each source, and only matching rows are streamed back to be joined in memory.
Passwords, SSH key passphrases, and tokens are saved to the macOS Keychain through the system Security framework — encrypted at rest and never written to disk in plain text. Connection settings live locally on your Mac.
None. Arris is local-first and privacy-first — there is no telemetry, analytics, or usage tracking. The only data that ever leaves your machine is a crash report, and only when you choose to submit one with an error report.
Yes. Arris understands dbt and SQLMesh projects natively — browse model lineage, run and test models, preview compiled SQL, and inspect plans without leaving the app.