n8n course for businesses and professionals: automations and AI agents

What you learn

How to build your first n8n workflows: the trigger, the connections to Gmail or Google Sheets, an AI step with OpenAI or Claude, and what to do when something fails. No code needed to get started.

What it is like

In an open group with other people who come on their own, or in private lessons; live online or in person in Catalonia.

Certificate

At the end you receive an IAescola certificate of participation.

Price

Open group course: €160 per person. Private lessons: €80 per hour or 5-hour pack: €350. VAT included.

Payment

Direct payment by bank transfer. FUNDAE funding only applies to companies in Spain.

1. 01Triggers and webhooks

   How a workflow starts: a form, an incoming email, a webhook, a spreadsheet that changes or a schedule. You learn to follow the data from node to node and to test every step before activating it.

2. 02Connections to your applications

   Gmail, Google Sheets, Calendar, a CRM, a database or any service with an API. We practise the connections you actually use and how to give each workflow only the permissions it needs.

3. 03Expressions and data logic

   Conditions, loops, expressions to read and transform data from one node to the next, and the code node when needed. You don't need to code to start; you do need to understand how data moves so you don't lose it along the way.

4. 04AI nodes

   How to call an OpenAI or Claude model from a workflow to sort emails, extract data from a document or write a draft, with clear instructions and an output the rest of the workflow can use.

5. 05AI agents in n8n

   How to build an agent that decides which tool to use, what limits to set on it and where to add a human review step when a mistake would be costly.

6. 06Error handling

   What happens when a workflow fails: alerts, retries and execution logs. We practise with missing data, applications that don't respond and formats that change.

7. 07Credentials and security

   How access keys are stored without leaving them written in nodes or in screenshots, who can reach the instance and which data should not be sent to an external model.

8. 08n8n cloud or self-hosted

   What it means to run n8n on your own server (installation, backups, updates) compared with using n8n's cloud service, and how to decide based on your data and your technical team.

• Administration, operations and marketing teams

  People without a technical profile who repeat the same tasks every week between email, spreadsheets and the CRM, and want to automate them without depending on someone else.

• Technical profiles who want a method

  Anyone who has already tried n8n on their own and wants to put what they know in order: expressions, error handling, credentials and deployment.

• People who have to decide

  Whoever must choose between cloud and self-hosting, or between n8n, Make and Zapier, and needs security and maintenance criteria before committing the team.

• Workflows of your own, running

  One complete workflow, from trigger to result, built with your applications and tested on real data.

• Judgement on using AI

  You know where to put an AI node or an agent, where a human review is worth it and which data should not leave for an external model.

• A reasoned hosting decision

  n8n cloud or your own server, with what each option means for maintenance, backups and control of your data.

• Workflows someone else can maintain

  Named workflows, a note on what they do, error alerts and properly stored keys, so the team doesn't depend on a third party for every change.

• A workflow that works is not a workflow that lasts

  Tutorials teach the happy path. Here we practise what happens when a value is missing, when an application doesn't respond or when a format changes, which decides whether a workflow is useful day to day.

• AI with oversight, not blind

  An AI node is sometimes wrong. You learn where a human review belongs, which data should not be sent to an external model and how to keep a record of what the workflow has done.

• Cases from your sector

  An accountancy firm, a shop, a clinic or a school do not automate the same things. The exercises start from your team's processes: invoices, customer emails, onboarding, periodic reports.

• Your team keeps the workflow

  The goal is that, when we finish, someone on your team can maintain and extend what has been built, with workflows named and documented, and not depend on a third party for every change.

1. 01

   We pick a real process

   Before starting we look at which repetitive tasks cost you most and which are good candidates: frequent, with clear rules and with data already in some application.

2. 02

   We build the first workflow together

   With your applications connected, we build a complete workflow step by step, from trigger to result, and test it with real data.

3. 03

   We add AI with control

   We add an AI node where it brings value, with clear instructions, a review step and a structured output the rest of the workflow can use.

4. 04

   Errors, credentials and go-live

   We leave the workflow with error alerts, keys stored properly, a name that makes sense and a note on what it does, so it can be activated and maintained with confidence.

Other courses you may be interested in