​

AI Support in Meetings

Topicflow AI can help before, during, and after meetings by analyzing work context, suggesting topics, and summarizing conversations. But humans run the meetings and make the decisions.

​

Before the meeting: Preparation

Topicflow AI can help you prepare by analyzing recent activity and identifying topics worth discussing. Suggested talking points Topicflow AI can suggest agenda items based on:

• Recent work activity (commits, PRs, completed tasks)
• Stale or off-track goals
• Overdue or incomplete action items
• Recent feedback given or received
• Upcoming deadlines or review cycles

Example prompts for preparation:

• “What has [person] been working on recently?”
• “What should I discuss in my next one-on-one with [person]?”
• “Are there any stale goals or overdue action items?”
• “What feedback has [person] received this quarter?”
• “What blockers or challenges have come up recently?”

What AI provides:

• A summary of recent work context
• Suggested discussion topics
• References to past meeting notes or action items

What AI doesn’t do:

• Decide what’s important (you do)
• Know about non-work context (personal situations, team dynamics)
• Replace human judgment about what to prioritize

Use AI suggestions as a starting point, not a script.

​

During the meeting: Assistance

AI can help while the meeting is happening, but it doesn’t run the meeting. Editing topic notes Topicflow AI can update meeting topic notes directly when you ask it to — for example, adding a summary, restructuring bullet points, or appending follow-up details. When AI writes or replaces content in topic notes, the output is automatically formatted as rich text (headings, lists, bold, etc.) so the notes stay clean and readable. The changes appear in the shared meeting notes so both participants see them. Extracting action items from notes As you take notes, AI can:

• Identify potential action items from your notes
• Suggest owners and due dates
• Convert informal commitments into trackable tasks

Summarizing key themes AI can highlight:

• Main discussion topics
• Decisions made
• Recurring themes across recent meetings

Formatting and cleanup AI can help:

• Clean up messy notes
• Add structure to free-form text
• Identify missing details (like action item owners)

What AI doesn’t do during meetings:

• Participate in the conversation
• Make decisions about what to discuss
• Replace human note-taking entirely

The conversation happens between people. AI assists with documentation.

​

After the meeting: Follow-up

Topicflow AI can help with post-meeting tasks. Recapping the meeting Topicflow AI generates a concise summary of each meeting as succinct bullet points — one per key topic or decision. Summary sections can be dismissed if they aren’t needed for a particular meeting. AI can also surface:

• A list of action items created
• Topics that were tabled for next time

Identifying follow-ups AI can suggest:

• Who should be notified about decisions made
• Whether any goals need updating based on the conversation
• If formal feedback should be created from informal comments

Connecting to past meetings AI can reference:

• Whether this topic has come up before
• How it relates to previous discussions
• Patterns over multiple meetings

Example prompts after a meeting:

• “Summarize what we discussed in today’s one-on-one”
• “What action items were created this week?”
• “Has [topic] come up in previous meetings?”

​

Meeting recordings and transcripts

When a meeting is recorded, Topicflow produces a transcript alongside the video. Click any line in the transcript to jump the video to that moment, and the URL updates as you move through the recording so you can share a link that opens at the same tab and timestamp.

​

How AI uses work context

Topicflow AI has access to:

• Integrated work activity (commits, PRs, tasks, tickets, etc.)
• Past meeting notes and agendas
• Goals and their progress
• Feedback given and received
• Action items created and completed

This context allows AI to:

• Ground suggestions in real work, not generic advice
• Reference specific events and accomplishments
• Identify patterns across time

Example of context-aware suggestions: Instead of: “Discuss project progress” AI suggests: “Discuss the billing refactor — 8 commits this week, PR pending review” Instead of: “Check in on goals” AI suggests: “Goal ‘Ship feature X’ hasn’t been updated in 3 weeks — discuss status”

​

Privacy and AI access

What AI can see:

• Work activity from integrations (based on your organization’s settings)
• Meeting notes where you’re a participant
• Goals you own or manage
• Feedback you’ve given or received
• Action items assigned to you or your direct reports

What AI cannot see:

• Private feedback not visible to you
• Meetings you’re not a participant in
• Other people’s one-on-ones
• Confidential performance data you don’t have access to

AI respects the same visibility rules as the rest of Topicflow.

​

Best practices

Use AI for preparation, not replacement Let AI suggest topics, but decide what’s actually important based on your knowledge of the person and situation. Verify AI summaries AI-generated summaries are helpful, but always review them. Don’t blindly trust that they’re accurate. Don’t over-rely on suggested talking points If something important is happening (personal situation, team conflict, burnout), that takes priority over AI suggestions. Let the conversation be human AI helps with preparation and follow-up, but the actual one-on-one should be a human conversation, not a checklist of AI-generated topics. Use AI to save time, not to avoid thinking AI can recall context and suggest topics faster than manual review. Use that saved time to think about coaching, not to skip preparation entirely.

​

What’s next