This is one of the best use-cases for Generative AI
A common scenario: Your IT talent lands a new job. They hand in their notice. Nobody has time to write documentation. The clock is ticking. How will you capture their tribal knowledge before they walk out the door?
The best way to preserve ephemeral team knowledge is with a series of Knowledge Transfer sessions (KT) so that the Subject Matter Expert can share their expertise with a group.
For most IT Engineers, when repetitive tasks are second nature, the steps become automatic, and the details fade into obscurity. These details may be forgotten and never communicated when a seasoned employees exits the organization.
By recording KT sessions on Zoom, AI can be used for real-time transcription, capturing and organizing the discussions. The value of an automated transcription is that it frees participants from taking notes, allowing them to fully engage, or ask questions, or even multitask in the background without missing crucial information.
♨
This is the power of Generative AI, that it can reliably produce a structured and comprehensive training guide even from disjointed discussions.
But the real game-changer is using Otter AI as the transcriber. Recently I uncovered Otter's ability to add screen-grabs from Zoom presentations, so as presenters visit new screens, Otter AI can automatically capture screenshots and sync them to the discussion.
Here is a recipe for scaling your KT sessions with AI:
Review task history: Jira | ServiceNow | GitHub (PR's) ... wherever the team is tracking their work.
Identify the Subject Matter Expert's top 10 critical or frequent tasks, that only they handle.
Block out one hour daily on the calendar for a team Zoom meeting.
Invite Otter AI as a meeting participant on the Calendar invites.
(Otter AI will capture and transcribe each meeting in real-time)
Have a Project Manager or Analyst mediate the sessions, while the SME performs the demonstrations.
Make sure a couple team members attend to add context and questions.
For each process in review, have the SME begin with a general overview and then delve into specifics as participants raise questions.
The information doesn't need to be presented in precise order. The AI will organize and structure the content later, even when there are gaps or twists in the discussion.
Paste any links for supplemental info in the Zoom chat.
Define internal jargon as it arises. Focus on clarity so anyone can make sense of the material.
The AI output will be a comprehensive Standard Operating Procedure, outlined-formatted-illustrated.
Publish it to your Knowledge Base and it's ready for your next new-hire replacement.
Afterwards, you can have someone new to the procedure follow it step-by-step, to ensure the AI-generated guide is truly comprehensive. Follow up on anything that's missing or unclear before the Subject Matter Expert vanishes.
With a tight economy right now, most companies are looking to reduce headcount and find offshore options, and we need effective ways to capture and scale Knowledge Transfer. As a TPM, my role is to streamline project execution, and even though discussing staff departures can be uncomfortable, this is a critical process-improvement for the benefit of any technology enterprise.
What has been your experience automating knowledge capture?
Have you tried similar methods?
