Comments anchored to the conversation
Highlight any line in a meeting transcript and leave a note pinned to that exact moment. Private by default, public when you want to share, and the anchor survives meeting reprocessing.
Annotate any moment
Open a meeting’s Transcript tab, drag-select any text — a single phrase, a full sentence, or a span across multiple speaker turns — and a floating Add comment button appears next to your selection. Click it, type your note, choose visibility, and the comment is pinned to that exact moment of the conversation.
Each comment stores the utterance index, the character offsets, the start and end timestamps, and the literal selected text. Click the comment row in the right column and the video seeks to that timestamp; the transcript scrolls the anchor into view and brightens the highlight band. While the recording is playing, the comment whose anchor covers the current moment lights up automatically.
Single line or across speakers
One speaker turn
Your highlight starts and ends inside one speaker’s turn. The most common case. The anchor records that utterance plus the start/end character offsets within it, and the comment row shows a single timestamp.
Across multiple turns
Your highlight spans from one speaker’s turn into another’s. The transcript renders the highlight as a continuous band across all the affected rows, and the comment shows the full time range — start → end. Useful for capturing a back-and-forth or a multi-speaker decision moment.
Public or private — your call, per comment
Every comment has a visibility setting that you control with one click.
Private (the default) means only you, the meeting owner, ever see the comment. Use it freely for reactions, follow-up reminders to yourself, or sensitive observations you’d never want a recipient to read.
Public means the comment is eligible to appear on share links you create. Whether it actually shows up depends on a separate per-share Include Comments toggle, which is off by default — public visibility is necessary, but not sufficient. Both have to be true.
Toggle visibility per-comment with the lock/globe icon on the comment row. A toast confirms the flip; if a share link is already live, the change reflects on the next refresh of the share view.
Survives reprocessing
Reprocessing a meeting regenerates the transcript from scratch. Comments are not blown away.
When reprocess completes, Notemesh walks every comment on the meeting and tries to re-locate it in the new transcript. If the exact text the comment referenced still appears, the anchor is silently re-pointed at the new location — utterance index, character offsets, timestamps all updated. No badge, no warning, no manual intervention.
If the wording shifted — a sentence got merged, a phrase rewritten, a typo fixed — the comment keeps its original quote and earns a small amber Transcript changed pill so you know the surrounding context may have moved. Hover the pill for the full explanation. The comment body, author, timestamp, and visibility setting all stay intact.
Before the new transcript replaces the old, the original is saved as an archived snapshot — but only when comments exist on the meeting, so reprocesses on uncommented meetings stay zero-overhead. Your comment body always references the wording you originally saw, even if the surrounding transcript looks completely different now.
Read-only on share links
When you share a meeting with the Include Comments toggle on, recipients see public commentary in a Comments section below the transcript — read-only, with privacy guardrails baked in.
Recipients can click a comment to seek the video and scroll the transcript to the anchored line, just like the authenticated view. They can’t add new comments — the selection mechanic is disabled on share views entirely. They can’t edit, delete, or change visibility on existing ones either.
Privacy is enforced at multiple layers. Private comments are filtered out at the database query — they’re never sent to the share endpoint at all, regardless of any UI state. The visibility concept itself is hidden from share viewers; they don’t see lock/globe badges, because to them every comment they see is simply “a comment.” Author email addresses and account IDs are stripped server-side; only the display name is exposed.
Want to reverse course on a share? Open the share dialog again, uncheck Include Comments, click Update — comments disappear from the share on the next page load. There’s no caching layer, no propagation delay.
Frequently asked questions
Can I select text across multiple speakers?
Yes. Cross-utterance selections are fully supported. Highlight from one speaker's turn into another and the floating "Add comment" button appears the same way. The transcript renders the highlight as a continuous band across the affected speaker rows; the comment row shows both start and end timestamps.
Are private comments ever exposed via share links?
Never. The server filters comments by visibility: public when serving the share endpoint — private comments are never sent to a share view at all, regardless of the Include Comments toggle. Both must be true: the share has Include Comments on, AND the specific comment is set to public.
Will my comments survive if I reprocess the meeting?
Yes. Reprocessing never auto-deletes comments. Each comment is matched against the new transcript by its selected text. If the wording still appears, the anchor is silently re-pointed and no badge appears. If the text shifted (rewording, merged sentence), the comment is preserved with its original quote and earns a "Transcript changed" pill so you know the surrounding context may have moved. The original transcript at the time of commenting is archived as well.
Can people with the share link add their own comments?
No. Share views are fully read-only for comments. Recipients can click a comment to seek the video and scroll the transcript, but they can't add new comments, edit existing ones, or change visibility. The selection mechanic that triggers the "Add comment" button is intentionally disabled on share views.
What does the recipient see about who wrote each comment?
Only the author's display name. Email addresses, account IDs, and the visibility toggle itself are all hidden from share viewers — they don't even know the "private" concept exists, because to them every comment they see is just "a comment." This protects authors from accidentally revealing more via a shared link than they intended.
Can I reply to or thread comments?
Not yet. Comments are flat — each one anchors to a piece of transcript, has an author and a body, and that's it. Threaded discussion is on the roadmap but the design for it ("at what point does this stop being a comment thread and start being a chat?") is still open. Two collaborators commenting on the same moment will produce two separate comment rows for now.
Related features
Comments are different from Knowledge Base notes — comments anchor to a specific transcript moment, while notes are general writing scoped to a tag. Use whichever fits the thought.
Pin your reactions to the moment they happened
Stop losing context in scattered chat threads and email replies. Comment directly on the transcript — private for yourself, public for the team. Start free.
Try Notemesh Free →