Onboarding a new hire: a Project that answers for you

Bringing on a new person costs a team more than anyone admits. A new hire doesn't know where anything is, and asks — rightly, since otherwise they'd be guessing. But every question pulls someone else out of their work, and the answer is the same one it was last time. Most questions in the first few weeks have exactly one correct answer, and it's already written down somewhere — nobody just knows how to find it. That's exactly what an AI Project solves: uploaded documents plus instructions, which a new hire can query in their own words, anytime, without feeling like a bother.

A typical scenario

Tom is joining a twelve-person sales team. Manager Lucy knows from past onboardings how this goes: for the first three weeks, Tom asks roughly fifteen questions a day, and each one costs a colleague not just a minute to answer but another ten minutes to get back into focus. Across the team, that adds up to several hours a week — and Tom, meanwhile, has the uncomfortable feeling that he's constantly bothering people, so on some things he just guesses instead of asking, and gets them wrong.

This time, Lucy sets up a Project called “Sales Onboarding” before Tom starts. She uploads the order-processing procedure, the price list with discount rules, two sample proposals, an approval matrix, and an overview of who owns what. On day one, Tom asks “how do we create an order here?” and gets a step-by-step process, including who needs to approve anything over a certain threshold. He hasn't stopped asking colleagues things — he just asks about what isn't in the documents, which is exactly where a human conversation makes sense.

How to do it

1. Gather what already exists. Don't write anything new. Procedures, the wiki, proposal templates, the price list, FAQ threads from chat, meeting notes about processes, the org chart. Most companies have this scattered across a drive, Notion, and email — the first job is putting it all in one pile.
2. Strip out what a new hire shouldn't see. A key step, not a formality. Salaries, personnel files, trade secrets beyond the role's scope, client personal data — out. A Project should contain exactly the part of the documentation that anyone on the team would show a new hire without a second thought. And run it on a business account with contractual data protection, not a private free chat.
3. Set up the Project and upload the documents. A Project is persistent context: uploaded files and instructions that AI knows in every conversation within that project. More detail in the tip on projects and persistent context.
4. Write the project instructions. This is where the quality gets decided. Something like: “You're answering a new hire on the sales team. Answer briefly, step by step, and always say which document you're drawing from. If the answer isn't in the material, say so explicitly and suggest who to ask — don't guess.” That last sentence matters most: without it, AI will fill in a missing procedure on its own, and the new hire will never know.
5. Give the new hire a starter set of questions. Ten typical first-week questions, so they know to ask: how to create an order, who approves a discount, where the proposal template is, who to contact about a complaint. That establishes the Project as the first reflex.
6. Track the gaps. Keep a log of questions the Project couldn't answer. It's the cheapest audit of your documentation you'll ever get — and a ready-made list of what to fill in.

The best tools

• Claude — Projects with persistent context and custom instructions; good when you want answers that reference a source document.
• Notion — a natural home for the company wiki itself; the Project then sits on top of it as a search layer.
• Company storage (Google Drive, SharePoint) — where the documents usually already live; you just need to pick the right subset, not move the whole drive.
• Connectors (MCP) — an advanced option for later: instead of uploading files, AI connects directly to the wiki or drive, see MCP connectors.

What you get out of it

• Team time: hours of every colleague's time that would otherwise go to repeated explaining. An interruption costs more than the answer itself.
• A calmer new hire: they can ask even the thing they're embarrassed to ask for the fifth time. No one watches, no one sighs.
• Consistency: everyone gets the same version of the procedure. No more each senior colleague teaching new hires a slightly different way.
• Feedback on your documentation: the list of unanswered questions shows you exactly where your procedures have gaps.

Pro tip

A Project goes stale faster than you'd expect — a price list changes, a process moves elsewhere, and AI will state the outdated version with total confidence. Put a quarterly review on your calendar: fifteen minutes to go through the uploaded documents, replace anything outdated, and note the last-updated date in the instructions. And hold to the site's core principle: AI proposes, the person approves. A Project is a guide, not an approval authority — a new hire doesn't send a contract or a pricing exception based on a chat answer; it goes out only after a person signs off.