The example in the lesson is made up. Fernway Logistics is a made-up distribution company, and its three new hires starting Monday, Priya Nair, Tom Becker and Lena Ortiz, are made up too. The practice needs no n8n, OpenAI or Google account, nothing is billed and no workflow is published. This is our own lesson, not an official n8n course.
What does the onboarding tracker agent actually do?
You paste in the updates the way they arrive. In the lesson, IT says Priya's laptop is ready, but her manager hasn't picked a first-week buddy. Tom emails to ask whether he can bring his new-hire paperwork on day one, and puts his bank details for direct deposit in the body of the email. Nobody has said anything about Lena.
The agent turns that into five entries:
- Priya's laptop: Done, owner IT, no follow-up needed.
- Priya's first-week buddy: Waiting, owner the hiring manager.
- Tom's new-hire paperwork: Waiting, owner HR, reply with your day-one steps.
- Tom's direct deposit: Unclear, owner Payroll. The bank details are not logged; the follow-up asks him to use your payroll system's form.
- Lena, all items: Unclear, owner HR. Check her offer letter, laptop and buddy today.
Then you type "Yes, save those to the sheet" and it writes all five rows to the Onboarding Tracker spreadsheet. Three need action before Monday.
How is the agent built in n8n?
The lesson follows the same path you would take in your own n8n:
- Sign up and get a key. Start at n8n.io with Get started free. On platform.openai.com, add credit, then create a secret key and copy it.
- Add a chat trigger. Click Create workflow, then Add first step… and pick On chat message, so you can paste updates into a chat.
- Add the AI Agent. Click the + on the trigger, open AI and pick AI Agent.
- Give it its orders. Under Options, click + Add Option, choose System Message and paste the system prompt.
- Give it a model and memory. Add OpenAI Chat Model with your key and pick gpt-5-mini. Add Simple Memory.
- Give it the Sheets tool. Add Google Sheets Tool, connect your Google account, choose Append Row, pick the sheet and map each column.
- Test it. Click Open chat, paste the updates, read the entries, then ask it to save them.
The standing instruction is the system prompt, and the trigger is a chat trigger. If you are new to agents in n8n, the n8n AI agent tutorial explains the parts in more depth.
Try it free, step by step
Build an n8n AI Agent That Tracks Every New Hire's Onboarding. Build an AI agent in n8n for Fernway Logistics that turns scattered onboarding updates from IT, managers and new hires into one row per item in Google Sheets, with the status, who owns it and the… You do every step yourself in a practice copy of n8n, it checks your work as you go, and “Do it for me” finishes any step you get stuck on. Free, and nothing touches your real accounts.
Start the free lessonWhy does the agent keep bank details out of the sheet?
Because the system prompt tells it to. It says never to record bank details, ID numbers, Social Security numbers, dates of birth, or medical or immigration document details, and to write only whether something was received. So when Tom puts his bank details in an email, the row says they came by email and are not logged, and asks him to use the payroll form instead.
That rule is a backstop, not a reason to paste everything. The lesson's finish screen says to paste only what you need, follow your company's rules for new-hire paperwork, and read every row before you share the sheet. Personal details like these are PII, and the safest copy is the one that never leaves your HR system.
Key takeaways
- One update can hold several items. The agent writes one row per item, not one per message.
- A fixed status list (Done, Waiting, Unclear) and an Owner column make the sheet easy to filter and chase.
- A new hire with no updates still gets a row, so silence doesn't hide a bad first day.
- The prompt keeps bank, ID and medical details out, but paste only what the tracker needs.
- The agent gives no legal advice about hiring paperwork or benefits. Reply with your company's own steps.
Frequently asked questions
- What columns does the Onboarding Tracker sheet need?
Five, typed in row 1: New hire in A1, Item in B1, Status in C1, Owner in D1 and Follow-up in E1. They match exactly what the system prompt tells the agent to return. In the Sheets tool, you then click the ✦ button next to each one under Values to Send, so the model fills them in.
- What do the three statuses mean?
Done means the update says the item is finished. Waiting means someone still has to act, and the Owner column says who: IT, the hiring manager, HR, Payroll or the new hire. Unclear is for a vague update, or a new hire with no updates at all. The follow-up then says what to check.
- Which onboarding items does it track?
The prompt names a signed offer letter, equipment, system access, a first-week buddy and new-hire paperwork as the items every Fernway hire needs before day one. Its examples include Offer letter, Laptop, System access, First-week buddy, New-hire paperwork and Direct deposit. You can change the list in the prompt to match your own checklist.
- Will the agent tell a new hire which documents to bring?
No. The prompt says not to give legal advice about hiring paperwork or benefits. For Tom's question, the follow-up tells HR to reply with your company's day-one paperwork steps.
- Why does the agent need memory?
Without memory, every message starts from nothing. You could not say "save those", because the agent would not remember which items you meant. Simple Memory keeps the recent conversation inside n8n; the lesson leaves its defaults alone (it keeps the last 5 messages).
- Which model does the lesson use?
gpt-5-mini, through the OpenAI Chat Model node. The lesson calls it fast, cheap and plenty for sorting onboarding updates.
- What does it cost to run?
You pay OpenAI for each request you send, not a flat monthly fee. New n8n accounts also include free credits: pick Gateway credits on the model step instead of a key. Check n8n's and OpenAI's pricing pages for current plans.
- What do I need to build it for real?
An n8n account (the cloud trial or self-hosted), an OpenAI API key with a little credit, and a Google account for the sheet. The n8n docs cover each node in detail.
- Does it work with other HR tasks?
The same build works for any list you keep by hand: swap the system prompt and the five columns. For another HR lesson in the same course, try a reply to an employee's leave question, or see how to log job applicants with n8n.
- What goes wrong most often?
Column headers that don't match the prompt, the Sheets node left on reading rows instead of Append Row, and personal details pasted that never needed to be. See n8n onboarding tracker mistakes for the full list, and n8n Google Sheets errors for connection problems.
