n8n

Our verdict on n8n

n8n is the workflow tool we use at Ryz Labs for AI automations, and in October 2025 Jordan called it one of the leaders in the category. It connects to the tools we already use and lets you watch every step run. The catch is the learning curve, so we skip the blank canvas and have Claude write the first draft of every workflow.

Our evidence is limited to one episode of hands-on use and one guest comparison, so read this as an early review, not a long-term one.

What we built with n8n

• n8n sales outreach workflow (Russell 2000 CEOs) (October 2025): a scheduled workflow that researches Russell 2000 companies with Perplexity, pulls CEO contact emails from a contact data API, writes a structured, customized email with ChatGPT, logs everything to Google Sheets and saves and sends the email from Gmail. Jordan ran it live by hand in Episode 17; a full run takes a few minutes.
• Claude-generated n8n podcast guest outreach workflow (October 2025): Jordan asked Claude, live on the show, for a workflow that invites podcast hosts onto Built This Week. He imported the JSON into n8n and got the full structure in seconds. Credentials and connections were still left to finish.

Sam said that this Claude-first approach is how we built several workflows in a couple of weeks.

How we use n8n

1. Describe the workflow to Claude. Jordan's prompt was one sentence: build me an n8n workflow that automates outbound outreach to podcast hosts asking them to be on our podcast. He did not give it many more instructions.
2. Download the JSON and import it. In n8n, create a new workflow and import from file (Episode 17, at 16:59). The podcast workflow arrived with a schedule and manual trigger, lead fetching, filtering, data prep, a website fetch, email generation and parsing.
3. Wire up the last 10% yourself. Connect credentials and check each node. Sam warned this part "can be at times challenging".
4. Start from a list and a set of questions. The sales workflow gets the companies first, then the research questions and topics, then splits the questions out per company.
5. Chain research, contact data and a model. Perplexity does the research, a data API finds the CEO's email, and the OpenAI API turns it all into structured output: the customized email.
6. Add wait steps. Jordan built pauses into the workflow so it does not spam any of the APIs.
7. Log to a sheet and draft before sending. Results land in Google Sheets, and the email is saved as a Gmail draft before it ships.
8. Schedule it, but test by hand. The workflow runs every day; clicking "execute workflow" runs it on demand so you can watch the logs and see how many tokens each OpenAI call uses.

But we've found a way to use Claude to build out, you know, 90% of really any automation you could think of and then just kinda tie it all together at the end.

— Sam Nadler, Episode 17

Where n8n falls short

• The learning curve. Sam said that while n8n is theoretically low code or no code, getting started "is a bit daunting".
• The last mile is manual. Even with Claude writing the structure, connecting the pieces is on you.
• Enterprise structure. Breeze founder Karim Malhas said in Episode 42 (at 02:55) that n8n originally had only one project with everything inside it, which did not work at the big enterprise he used to work for. He built Breeze's workflow builder to look like n8n on purpose, but with projects, teams and segregation.
• Not much different from what came before. Jordan said n8n is probably not that different from Zapier and Make; its edge is the agentic workflow model that others are now copying.

We did not cover n8n pricing or self-hosting on the show.

n8n compared

• n8n vs OpenAI's agent workflow tool: Jordan tried OpenAI's version after using n8n and found it far less capable (Episode 17, at 16:26).
• n8n vs Breeze: for voice agents at enterprise scale, Karim argues Breeze adds the multi-project setup n8n lacked.

It's very, very rudimentary, I would say, now. It definitely doesn't have as many features that n n n eight n does.

— Jordan Metzner on OpenAI's agent workflow tool, Episode 17

Episodes featuring n8n

• Episode 17: the Russell 2000 sales outreach workflow running live, and a podcast outreach workflow generated by Claude.
• Episode 42: Karim Malhas on why Breeze is "VAPI meets n8n for enterprises".

FAQ

What is n8n used for?

Jordan describes n8n as a platform for agentic workflows, or workflows in general. We use it for sales outreach: researching companies, finding contacts, writing a custom email with ChatGPT and sending it from Gmail.

Is n8n hard to learn?

It is sold as low code or no code, but Sam said there is a learning curve and it is a bit daunting at first. Our shortcut is to ask Claude to generate the workflow as JSON and import it.

Can Claude build an n8n workflow?

Yes. In Episode 17 Jordan prompted Claude for a podcast guest outreach workflow, downloaded the JSON and imported it into n8n in seconds. The credentials and connections still needed finishing, which Sam called the last 10%.

Is n8n better than OpenAI's agent workflow tool?

In October 2025 Jordan said OpenAI's agent workflow tool was very rudimentary and had far fewer features than n8n. We have not retested it since.