Automatisch

What Automatisch is

Automatisch is an open source business automation tool that, in the home page's words, "helps you to automate your business processes without coding." It connects the online services a team already uses, such as Slack, Google Sheets or Stripe, so that work moves between them without code. The docs say you do not need any programming knowledge to use it. You can run it on the company's cloud solution, which the site calls affordable, or self-host it on your own servers. The site says it is made in Berlin.

The reader it fits is an operations person, a small business or an IT team that wants app-to-app automations but needs to keep the data and the runtime under its own control. The home page describes the people behind it as "software engineers who see the power of automation and love solving complex problems," whose aim is to help businesses automate their workflows "without investing a lot of time and effort." It is an automation product only: there is no page or app builder in it, and the output of a workflow is data moving between services.

What you build and how

You build a flow out of steps. The documentation's worked example has two: search all tweets that include the Automatisch keyword, then post those tweets to a chosen Slack channel. The company says this is one of the internal workflows it uses to test the product. A flow is prepared once and then runs continuously until you stop it or the connected account is unlinked. The docs put the current run interval at every 15 minutes, and say the team plans to extend that and to support instant updates where a third-party service offers them.

The app list is the real inventory of what you can connect. The docs list these as currently supported: Airbrake, Airtable, Anthropic, Appwrite, Better Stack, Brave Search, Carbone, Changedetection, ClickUp, Cryptography, Datastore, DeepL, Delay, Discord, Disqus, Dropbox, Eventbrite, Filter, Firefly III, Flickr, Formatter, Frappe and ERPNext, FreeScout, Ghost, Gitea, GitHub, GitLab, Gmail, Google Calendar, Google Docs, Google Drive, Google Forms, Google Sheets, Google Tasks, HTTP Request, HubSpot, Invoice Ninja, Jotform, LibreTranslate, Mailchimp, MailerLite, Mattermost, Miro, Mistral AI, Monday, Notion, Ntfy, Odoo, OpenAI, OpenRouter, PDFMonkey, Perplexity, Pipedrive, Placetel, PostgreSQL, Pushover, Reddit, Remove.bg, RSS, Salesforce, Scheduler, SignalWire, Slack, SMTP, Spotify, Strava, Stripe, SurveyMonkey, Telegram, Todoist, Together AI, Trello, Twilio, Twitter, Typeform, VirtualQ, Vtiger CRM, Webhooks, WordPress, Xero, You Need A Budget, Youtube and Zendesk. Several entries are utilities rather than apps, such as Filter, Formatter, Delay, HTTP Request and Webhooks, so a flow can filter, reformat and delay data and call a service that has an HTTP endpoint. A page for requesting an integration exists, and the docs include a developer section on building your own: folder structure, the app and global variable files, auth, triggers, actions and examples.

The AI features that appear on the pages are connectors. Anthropic, OpenAI, Mistral AI, OpenRouter, Perplexity, Together AI, DeepL and LibreTranslate all appear in the app list, so a flow can send text to a model or a translator as one step.

Self-hosting is documented in detail. The install guide covers Docker Compose and plain Docker; if you need another install type you open a GitHub issue. With Docker Compose you clone the repository and run a single command, then browse to http://localhost:3000. With plain Docker there are two services, main and worker, that use the same image and the same environment variables, except that the worker has WORKER set to true. Upgrading means pulling the main branch and rebuilding with docker compose up --force-recreate --build.

The configuration reference lists the settings you manage. Host, protocol and port default to localhost, http and 3000. The database is PostgreSQL, with host, port, schema and SSL settings, and there are Redis settings too. A BullMQ dashboard can be turned on with its own username and password. SMTP settings are included, and telemetry is on by default and can be disabled. Two keys deserve care: ENCRYPTION_KEY encrypts the credentials for third-party services and WEBHOOK_SECRET_KEY verifies webhook requests. The docs warn that if you change either one, your existing connections and flows will stop working.

Licensing, community and claims

The home page gives three reasons to choose it. Data privacy: you can store your data on your own servers, which the site says matters for companies in sectors like healthcare and finance and for European companies bound by the General Data Protection Regulation. Open source: anyone can influence how it is developed. No vendor lock-in: you keep control and avoid dependence on one vendor or technology. The docs spell that last point out: if Automatisch stops being useful, switching to another provider is easier "since you have all data and flexibility."

The license is AGPL-3.0, stated in the contribution guide, which says contributions are licensed under it. Contributing follows a plain GitHub process: fork the repository, branch from main, update the docs, run yarn lint, open a pull request, and use conventional commit messages. Bugs are reported as GitHub issues. The community lives on GitHub, Discord and Twitter, and the home page asks visitors to star the project. The docs are organized into Getting Started (what it is, installation, key concepts, creating a flow), Integrations, Advanced settings (configuration, credentials, telemetry), Contributing and a section on building integrations, plus pages on the license and the community. The docs ask readers who find a gap or have a suggestion to file an issue or send a tweet to the project's Twitter account.

The cloud option is called affordable; the site publishes no prices for it. The home page shows a row of company logos under "Trusted by these companies." The privacy argument rests on self-hosting: the data stays on servers you run.

Who it suits and who it does not

Automatisch suits a technical operations team, an agency or a company in a regulated field that wants a no-code automation layer on infrastructure it controls. Someone who can run Docker and keep a PostgreSQL database and the two secret keys safe gets a no-code automation tool under an open license. It also suits developers who want to write their own integration and contribute it back.

It is a weaker fit for someone who wants a hosted tool with a long list of ready-made templates and nothing to install. The 15-minute run interval rules out use cases that need an instant reaction to an event. The app list is the one to check: if a service you need is missing, the options are the HTTP Request and Webhooks steps, a request through the docs, or building the integration yourself. Because AGPL-3.0 applies, a company planning to modify and redistribute it should read the license first.