Building Onboarding Ecosystem

With significant staff turnover approaching, much of the team's operational knowledge existed only in people's heads. Existing onboarding documentation was scarce and outdated, creating a real risk of losing critical know-how during the transition.

Without action, new employees would arrive to a scattered handover, and remaining colleagues would get swamped with routine questions — slowing everyone down at the moment the team could least afford it.

1. 01Interviewed each role holder to capture what they knew but had never written down — recurring tasks, tricky decisions, and the unspoken shortcuts that keep things running.
2. 02Wrote clear SOPs from those interviews, using a consistent format so newcomers could find what they needed without asking for help.
3. 03Designed self-directed onboarding plans that mapped out the first weeks, key contacts, and essential tasks for each role.
4. 04Recorded screen walkthroughs for recurring workflows where written steps alone would leave too much room for interpretation.
5. 05Built a Custom GPT trained on the SOPs and internal materials, so colleagues and customers could get answers on their own without waiting for a teammate.

• Knowledge capture over knowledge transfer: interviews extracted what people knew but had never written down, so departing colleagues weren't the only ones who knew how things worked.
• Self-directed by design: role-specific onboarding plans let new employees get up to speed without constant hand-holding from a shrinking team.
• Different formats for different people: SOPs, screen recordings, and an AI assistant covered reading, watching, and asking — whichever fit the moment.
• AI as on-the-job help: The Custom GPT turned static documentation into an interactive resource, so team members and customers could find answers on their own and stop pinging busy colleagues with routine questions.