← sources

ByteChef: open-source workflow automation platform (1k+ GitHub stars, 150+ forks)

source: reddit (reddit:selfhosted)published 2026-10-20 07:14ingested 2026-10-21 02:38

retained text

Hi all, we are a small team from Croatia who are behind ByteChef, an open-source workflow automation and AI agent platform, self-hosted alternative to n8n / Zapier / Make.

We have been working with low-code automation platforms for the better part of the last 10 years, and we always found something lacking. So, in our free time, we started building our own automation platform, which I'm happy to say we now use extensively in our everyday work.

It started as a JHipster project (who even remembers those code generators in the age of AI 😭), and today it is a fully fledged automation platform with AI capabilities. Here is a short list of features, just so you get a better picture of what ByteChef does:

* Visual workflow editor: drag-and-drop actions with conditions, loops, branching, sub-workflows, and error handling.
* 250+ built-in integrations: CRMs, Slack/Discord, Google stuff, databases, cloud storage, e-commerce + a plain HTTP component for anything we don't cover yet.
* Code steps: write JavaScript, Python, Ruby or Java right in a step when no-code action isn't enough.
* AI agents as workflow steps: you drop in an AI Agent component and give it a model, tools, memory, and guardrails. It works with Ollama if you want everything to stay local.
* Knowledge bases / RAG
* MCP in both directions: you can use MCP servers as tools, and you can also expose your own workflows as MCP tools to other AI clients.
* Copilot: you describe a workflow in a sentence, and it builds a first version, helps with debugging workflows, etc.
* Built-in data tables, so simple workflows don't need a separate database.
* Write your own connectors: there's a CLI that scaffolds a component from an OpenAPI spec.
* GitHub: [https://github.com/bytechefhq/bytechef](https://github.com/bytechefhq/bytechef)
* Docs: [https://docs.bytechef.io](https://docs.bytechef.io)

Try with one command, no infra to set up (data lives in \~/.bytechef):

    docker run --name bytechef -it -p 8080:8080 \
        --env BYTECHEF_DATABASE=h2 \
        --env BYTECHEF_AI_MEMORY_PROVIDER=in_memory \
        -v ~/.bytechef:/root/.bytechef \
        docker.bytechef.io/bytechef/bytechef:latest

If you plan to keep it, use the Docker Compose setup with Postgres. The knowledge base and other AI capabilities like RAG need pgvector, and there's no migration path from H2. Helm chart is in the repo too!

What would stop you from running this in your homelab? We would love your feedback, as honest answers help us decide what to fix next.

document fetched 2026-10-21 00:29 · canonical https://reddit.com/r/selfhosted/comments/1xalgp5/bytechef_opensource_workflow_automation_platform