People & HR

How to offboard a departing employee with AI

People & HR3 AI tools7 steps6 friction points

Want an agent to handle this on your live data? See the Starch version →

Offboarding a departing employee is one of those workflows that looks simple on paper and turns into a multi-day coordination problem in practice. You're revoking access across a dozen tools, recovering or reassigning files and passwords, scheduling knowledge transfer sessions, notifying stakeholders, processing final pay, and making sure nothing falls through the cracks — all while someone is mentally checked out and you have other things to run.

It feels like an AI problem because so much of it is documentation, drafting, and checklist logic — exactly what language models are good at. You need an offboarding email to the departing employee, a checklist for IT, a handoff brief for the team picking up their projects, and a set of calendar invites for knowledge transfer sessions. None of that requires human judgment. It requires someone to generate structured, accurate text quickly.

ChatGPT, Claude, and Gemini can genuinely help here. You can use them to draft every communication in the offboarding sequence, build a tool-by-tool access revocation checklist, summarize the employee's open projects into a handoff doc, and produce a final-day schedule. The quality is good. The gap is that none of it connects to your actual systems — you do all the moving.

Skip the manual work → See the Starch version

People & HR3 AI tools7 steps6 friction points

AI walkthrough

How to do it with AI today

A practical walkthrough using ChatGPT, Claude, and other off-the-shelf LLMs — what they're good at, what you'll have to do by hand.

Tools that work for this

ChatGPTClaudeGemini

Step-by-step

1 Start with Claude or ChatGPT and paste in the employee's name, role, start date, last day, and a bullet list of the tools they have access to. Ask the model to produce a master offboarding checklist organized by owner: IT, HR, manager, and finance.

2 Take your list of the employee's current open projects — paste in whatever you have from Notion, Asana, or email — and ask the model to turn it into a one-page handoff brief with context, next steps, and a suggested assignee for each item.

3 Draft the departure announcement to the team. Paste in the employee's role, tenure, key contributions, and the replacement plan. Ask Claude to write a 150-word internal Slack message that's warm but doesn't over-explain.

4 Generate the outgoing email to the departing employee that outlines their final-week schedule, what they need to hand off, and what to expect on their last day (equipment return, final paycheck timing, benefits end date). Paste in your actual company policies if you have them in a doc.

5 Use ChatGPT or Gemini to draft a list of knowledge transfer meeting topics — one per project or system they own — so you have an agenda ready for each session. You can paste these directly into calendar invites.

6 Ask the model to draft the IT deprovisioning ticket: every app, every access level, in a format your IT person can work through sequentially. Be explicit about the apps — the model won't know your stack unless you list them.

7 Once you have all the documents, manually send each one, create calendar invites, submit the IT ticket, and file the HR paperwork. Everything the LLM produced needs to be copy-pasted and acted on by you.

Prompts you can copy

Generate a complete offboarding checklist for a departing marketing manager whose last day is Friday. They have access to: Google Workspace, Slack, HubSpot, Notion, Canva, LinkedIn Ads, and our AWS account. Organize by owner: IT, HR, manager, and finance.

Turn these project notes into a handoff brief: [paste notes]. For each project, summarize the current status, the next three actions required, any blockers, and who on the team would be best to take it over.

Write an internal Slack announcement for a departing employee. Name: Jordan. Role: Head of Growth. Tenure: 2 years. Last day: June 13. Keep it under 150 words, warm but professional, and do not mention why they're leaving.

Draft a final-week schedule email to a departing employee. Include: daily knowledge transfer sessions with their team lead, equipment return instructions for Friday, and a note that their final paycheck and benefits end date are handled by HR.

Write an IT deprovisioning ticket for an employee whose last day is this Friday. List every step required to revoke access to: Google Workspace, Slack, HubSpot, Notion, GitHub, AWS, and Gusto. Flag anything that requires a 24-hour lead time.

Reality check

Where this gets hard

The walkthrough above works — until your numbers change, the LLM hallucinates, or you have to re-paste everything next month.

No connection to your actual tool stack — you have to manually list every app the employee touches, and if you forget one, it stays open until someone notices.

The checklist is a document, not a workflow. Nothing in it gets tracked, assigned, or marked complete unless you build that separately in another tool.

Context drops between sessions. The handoff brief you drafted Monday and the access revocation list you drafted Wednesday exist in different chat windows with no shared memory.

Nothing triggers automatically. Every offboarding starts the same way: you remember to open the LLM, you paste in the context again, and you chase down completion manually.

Outputs aren't connected to your HR system, calendar, or email. Each document has to be copy-pasted, reformatted, and sent by hand — which is where things get dropped.

No audit trail. Once the chat is closed, there's no record of what was done, by whom, or when — which matters when an auditor or a departing employee's lawyer asks later.

Tired of the friction?

Starch runs the whole workflow on live data — no copy-paste, no hallucinated numbers, no re-prompting next month.

See the Starch version →

Starch alternative

The same workflow on Starch

Starch is an agentic operating system. For employee offboarding, that means describing the workflow once in plain English and having an agent build a persistent app that runs the checklist, drafts the communications, and tracks completion against your live calendar and email data — without you re-prompting it every time.

Connect Gmail and Google Calendar through Starch's scheduled sync; the agent can read active threads with the departing employee, surface unresolved items, and draft handoff communications using real context from your inbox — not a blank prompt.

Describe the offboarding app you want: 'Build me an offboarding tracker that creates a checklist per employee, assigns tasks to IT and HR with due dates, and marks items complete.' Starch builds it. It persists and runs for every future departure.

Starch's Knowledge Management app keeps offboarding docs, process notes, and role-specific handoff templates in one searchable place — so the next time someone leaves, the institutional knowledge doesn't leave with them.

Use the Task Manager app to track every open offboarding action — IT ticket, benefits processing, equipment return, final paycheck — with P1–P4 priority levels and overdue alerts, so nothing sits incomplete after day one.

Connect your HR tools (Gusto, BambooHR, Rippling, Paylocity, ADP, and others are reachable through Starch's integration catalog or scheduled sync) so the agent can pull the departing employee's data into the offboarding workflow without you copying it over.

Starch automates the calendar side through browser automation — no API required. Describe a prompt like 'Schedule 30-minute knowledge transfer sessions for each of Jordan's five active projects in the last week of their tenure' and the agent builds the scheduling logic into your workflow.

Get closed-beta access →

Toolkit