Build an AI Maintenance and IT Request Form

Broken dock doors, dead scanners and flickering lights usually get reported by shouting across the floor or sending a text to whoever seems to be in charge. In this free lesson you build a real internal request form for Ridgeline Supply Co. in Lovable, then connect it to an n8n AI agent that sets the priority, assigns a team, saves the request to Google Sheets and emails the person a confirmation.

What you will build: a maintenance request form with an AI backend

The front end is a simple web form built in Lovable from a written prompt, with no coding. It is titled "Report a Problem" and asks for four things: name, email, what needs fixing, and location. When someone submits it, the form sends the details to an n8n webhook.

The back end is an n8n workflow: a webhook trigger, an AI Agent with an OpenAI chat model, a Google Sheets tool, a Gmail tool, and a Respond to Webhook node that sends the form its confirmation. The agent reads each request and sets a priority: Urgent if it is a safety risk or stops shipping or receiving, This week, or Backlog. It assigns one team: Facilities, IT, or Equipment. Then it adds the request to a spreadsheet called Requests and emails the requester the priority and team.

In the test, Jordan reports that dock door 3 will not open with two trucks waiting. The agent treats it as urgent for Facilities.

Who this is for

This lesson is for operations managers, facilities leads and office managers at small businesses without a help desk. If requests arrive by text, sticky note and hallway conversation, and nobody knows what is waiting or who owns it, this gives you one front door and a log. You do not need to code. You will create accounts on Lovable, n8n, OpenAI and Google to build it for real.

Step by step: what you do in the lesson

• Sign up: start in Lovable and click Get started for free.
• Build the form: paste the prompt for the maintenance and IT request form and click Build.
• New workflow: switch to n8n and create a workflow.
• Webhook trigger: add "On webhook call" and set it to respond using a Respond to Webhook node.
• AI Agent: add the agent, set the prompt source to Define below, paste the prompt template that pulls name, email, request and location from the webhook, and paste the system message.
• Chat model: add the OpenAI Chat Model with your API key and pick gpt-5-mini.
• Sheets tool: add the Google Sheets Tool, then create a spreadsheet named Requests with the headers Name, Email, Request and Location.
• Connect the sheet: sign in with Google, pick the Requests document and Sheet1, and let the model fill all four columns.
• Gmail tool: add the Gmail Tool and sign in with Google so the agent can send confirmations.
• Respond node: add Respond to Webhook after the agent.
• Publish: make the workflow live.
• Add the secret: copy the webhook production URL and save it in Lovable as the N8N_WEBHOOK_URL secret.
• Test the form: preview the app, fill it with test data, click Submit Request, and check the new row in Google Sheets.

How the AI decides priority and team

The system message is only a few lines, and that is on purpose. It names the business, defines the three priorities and the three teams, and gives one clear rule for Urgent: anything that is a safety risk or stops shipping or receiving. It tells the agent to use the Google Sheets tool to add the request and the Gmail tool to send a short confirmation with the priority and team. The spreadsheet row holds exactly what the person typed. The priority and team go into the confirmation email.

Common problems this avoids

• Lost requests. Every submission becomes a row, so nothing lives only in a text thread.
• No reply. The requester gets an email with the priority and team, so they stop asking if anyone saw it.
• Everything marked urgent. One written rule decides what is Urgent, instead of whoever shouts loudest.
• Wrong team. IT problems go to IT and a stuck door goes to Facilities from the start.

Why it works

The webhook keeps the form and the workflow separate, and the URL lives in a Lovable secret instead of the page code. The Respond to Webhook node means the person only sees "Request received!" after the agent has finished. Every request ends up in one sheet, in the order it came in, and the person who reported it gets an email instead of wondering if anyone heard.

How to use this in your operation

Rewrite the priorities and teams to match how your building works. You might add a Safety team or a rule that anything touching a forklift is Urgent. Put a QR code to the form at each dock, in the break room and at the front office. Review the Requests sheet at the start of each shift, and use it to spot repeat problems, like the same door failing every month. The AI sorts and confirms; your team still decides what gets fixed first.

Frequently asked questions

Can AI prioritize maintenance requests?

It can make a first pass. In this lesson the agent follows a short written rule to mark each request Urgent, This week or Backlog and to assign Facilities, IT or Equipment. A person should still check the list, especially anything marked Urgent.

Is this lesson free?

Yes, Stepthrough is completely free. To build it for real you need your own Lovable, n8n, OpenAI and Google accounts. OpenAI charges per request, and Lovable and n8n have their own plans, so check their current pricing.

Do I need to write any code?

No. Lovable builds the form from a written prompt, and every n8n node is added by clicking. The lesson is 87 short steps and about 15 minutes.

Where do the priority and team show up?

In the confirmation email the agent sends through Gmail. The Google Sheets row keeps the name, email, request and location exactly as they were submitted.

How does the interactive lesson work?

Each step is shown to you first on the left (Watch). Then you do it yourself in a practice copy on the right (Your turn), and every click is checked. If you get stuck, press Do it for me. It is a practice copy, so no real accounts are touched and no emails are sent.

More in AI Automation for Operations