Importing an existing n8n workflow and letting GENI run it
Already built workflows in n8n? Then you do not need to rebuild anything. Import the JSON, and GENI runs the steps from your conversation, without an n8n server of your own.
New to this topic? Start with How to Build Your First Agentic Workflow with dGENIX
Who this is for
n8n is a popular way to automate processes with blocks you connect on a canvas. Many business owners and agencies have built workflows in it: enriching leads, moving data from one system to another, collecting reports.
Those workflows work, but they run somewhere. On your own server that needs maintaining, or in an environment you pay for separately. And to start one, you have to open n8n.
With the Workflow Builder in dGENIX you import such a workflow and let GENI run it. You start it with a sentence in your conversation, and you do not need an n8n server of your own for it.
How it works
The idea is simple. You build and change the workflow where you are used to doing that: in n8n. The running happens in dGENIX.
| What | Where |
|---|---|
| Building and changing | In n8n, on the canvas |
| Importing | In dGENIX, by pasting or uploading the JSON |
| Running | In dGENIX, with a sentence to GENI |
| Connections and credentials | From your dGENIX connectors, not from the JSON |
GENI carries out each step with the tools built into the platform, from web requests to email and image generation. Afterwards you get back what happened in each step.
Step 1: export the workflow from n8n
Open the workflow in n8n and export it as a JSON file. Export the whole workflow, not a selection of a few blocks. An export of part of the canvas is the most common reason an import fails.
Step 2: import into dGENIX
Activate the Workflow Builder skill from the skills in your dashboard. Then paste the JSON into the chat, or import the file through the dashboard.
GENI shows which steps it recognised. For example, for a workflow that enriches leads: a webhook that receives a company name, a request that fetches company details, a filter, a step that fetches the website, an AI step that summarises a description, and a step that adds a row to Google Sheets.
Check that list. If a step from your workflow is missing, it is not supported.
Step 3: check the connections
A workflow from n8n uses credentials stored in n8n. They do not come along in the export, and that is a good thing. In dGENIX they come from your own connectors.
If your workflow uses Gmail or Google Sheets, that connection needs to be active in dGENIX. Do that before you run the workflow for the first time. A workflow that runs but does nothing is almost always missing a connection.
Step 4: run it
Just ask:
"Run the lead enrichment for De Krent Bakery."
GENI carries out the steps and reports the result: what was found, where it was saved, and whether a step gave a warning. If something looks odd, ask what happened in each step.
Step 5: schedule it
If the workflow needs to run at fixed moments, combine it with a scheduled task. For example, every Monday at eight, the workflow that enriches last week's new requests. How scheduling works is covered in scheduled tasks explained.
The limits you need to know
The Workflow Builder is not a full replacement for n8n. A few things to know in advance:
- Not every n8n node is supported. dGENIX carries out the steps itself. Nodes that require their own n8n environment do not work.
- There is no visual editor. You make changes in n8n, then import again.
- Changes in n8n do not come along automatically. If you adjust the workflow there, import it again.
- Slots apply per imported workflow, not per run. The Pro plan includes three slots. Delete a workflow you no longer use to make room.
- Each step is charged separately. A workflow with many steps or AI steps costs more than a short one.
Workflow Builder, Custom Workflows or a scheduled task?
dGENIX has three ways to let work run automatically. They look similar, but do different things:
| Way | What it does | When to choose it |
|---|---|---|
| Scheduled task | GENI carries out an instruction in plain language at a fixed moment | You can describe the task in a sentence |
| Workflow Builder | GENI runs an imported n8n chain step by step | You already built the workflow in n8n |
| Custom Workflows | GENI calls a webhook on your own server | The process has to keep running on your own systems |
They can be combined: a scheduled task that starts your imported workflow is exactly that kind of combination. Why an agent adds something next to classic automation is covered in AI agent or no-code.
What you gain
The biggest advantage is not in the technology, but in how you work. A workflow you start with a sentence gets used more often than one for which you have to open another system. And because GENI reports the result in plain language, you notice sooner when something goes wrong.
You can use the result straight away in the same conversation. Have the enriched leads assessed right away, for example, or let the SEO Engine review the websites from your list. How to build chains like that is covered in build your first agentic workflow.
Frequently asked questions
Do I still need an n8n server?
No. You only use n8n to build the workflow. The running happens in dGENIX.
Can GENI build an n8n workflow for me?
It can think along about the setup, but you build the JSON in n8n. Describe the process and ask which steps you need; then build it on the canvas.
What happens to steps that are not supported?
They are skipped. Ask GENI after importing which steps were recognised, then you see exactly where it goes wrong and can replace that step in n8n.
