Hey everyone 
If you’re anything like me, you spend literally all of a lot of your time in meetings.
The Grammarly team is testing a new Notetaker Agent and is looking for a small group from our community to try it early. Basically, it joins your Zoom or Google Meet calls (as an example) and turns the conversation into actual structured notes you can use.
If you’re running team meetings, client calls, project syncs, workshops — especially if you’re already moving that stuff into Coda afterward — this could save you a ton of time.
Fair warning, though: this isn’t just a quick preview. We need people who can actually commit to about 3 weeks of testing, which means:
-
Actually using it in your real meetings
-
Giving thoughtful feedback (not just “looks good”)
-
Jumping on a few feedback calls with the team
So yeah, this is really for people who want to help shape where this thing goes, not just kick the tires and ghost 
If that sounds like you, please fill out this form, and we’ll reach out early to mid next week with details. Questions? Hit me up here on Discourse or email me at ruggy.joesten@grammarly.com.
Some helpful context:
- You’ll need a Grammarly account since the Notetaker comes with the Grammarly/Superhuman desktop app.
- Make sure you’ve got the latest version of the app.
- You’ll find the Notetaker settings under Settings, where you can adjust:
- Auto-join for meetings (if you’ve connected Google Calendar)
- Notification preferences
- Calendar integration (Google Calendar is supported right now)
See below to check when making sure your Grammarly account is up to speed:
3 Likes
If you can compress time and steps while adding additional value, this might be a great addition for several roles and teams.
I’m a hard pass on this one only because my entire computer is a listener. Everything I see, say, hear - all of it is captured and parts relevant to individual Coda documents find their way into them without effort. But the real advantage is doing this with a local-first approach.
1 Like
Only Zoom and Google Meet?
This would be really helpful, but my company is an MS shop
1 Like
This is one of the reasons I use an OS-based approach. The OS is the right abstraction to gather content, so it doesn’t matter which app you use—you get 100% of the conversation’s audio, video, and transcripts. I use Pieces.
And another key observation—a notetaker for meetings only is the tip of the notetaking objective. You really want it on all the time, capturing noteworthy moments regardless of context throughout the workday. From that, you then want to create “notes”, possibly persisted in a variety of target apps, each with its own index and contexts.
The best notetaker is no notetaker.
1 Like
Does it actually „join“ the meeting as a bot or is it just listening via audio? Would like to test, but only if its option two.
1 Like
Good question! Totally fair to check this before you try it.
Nope, it doesn’t join as a visible bot or show up as an extra participant. Nothing in the attendee list.
It listens to the audio locally through the Grammarly/Superhuman app on your computer and generates notes afterward. So if you’re worried about a bot showing up in the participant list, you’re good.
Just a few things to know:
- You’ll need a Grammarly account since the Notetaker comes with the Grammarly/Superhuman desktop app.
- Make sure you’ve got the latest version of the app.
- You’ll find the Notetaker settings under Settings, where you can adjust:
- Auto-join for meetings (if you’ve connected Google Calendar)
- Notification preferences
- Calendar integration (Google Calendar is supported right now)
See the below screen shots to help make sure you have the right toggles on, and that you have the most updated version of the app on your device:
1 Like
My experience with notetakers is since the ML era and in the past 2 years with tools that pivoted from transcription to note-taking tools. After more than 12 years with audio notetakers I can say that there are some very important features that most tools miss. Unfortunately, I don’t use Grammarly (nor Superhuman), so I won’t be able to join this beta. But if there is some sort of video preview or else, maybe I could contribute with some practical usecases and insights.
My last research on the notetakers is from early 2025, so consider the below as likely outdated:
- Voice tagging - very important because I can say one word to tag specific part of a conversation and that helps a lot. The approach of tana .inc is really cool for using hashtags for sorting notes. If Coda wants to explore notetaking - just buy Tana and incorporate it so that tagging and coda tables are working seamlessly and we are all done with adding data to tables for the next decade.
Scouting software in the football (the foot-touching-ball-game 
) for auto-tagging video content is maybe a good case to consider.
- Requesting information with voice - kind of relates to the above
- Video notetaking - I prefer video notes so much more than just audio. Even for in person meetings when we use whiteboards, documents, write on screens, etc. and I want all this video context in the notes. Haven’t yet found a notetaker that can really understand the visual context for why about 1/15 notes I watch the video recording because need the visual context
- Watching my screen on mobile and desktop - gemini live kind of thing. I know some people are already there with some self-made tools but I am not that tech-savvy
- Multiple-ends notetaking - in EU recording mobile phone calls is (almost) impossible but if there is an app which can listens to both sides of a phone call would be super useful; alternative a quick “join the call link” which allows the mic recording… I am asking for too much
- Limited language support - English is far from being universal and with the decreasing need of knowing languages (LLMs effect) it happens more often that we have great team members who don’t particularly know English
- Multilingualism - it’s not uncommon to have conversations with multilingual people and switching between languages happens more often lately in meetings
- Mobile one-button Widgets - never thought that 5 seconds of opening an app and clicking a recording button would be so annoying and cut conversations so badly like “hold on, give me sec to turn the notetaker on”.
- Pausing button - not always present
- Export formats are limited - need to be able to export - video, audio, summary, video with subtitles, etc. One of the reasons I fancy Happyscribe because it can export the content in all those video/subtitling production formats which are quite handy.
- “Avatar meeting presence” - not missing anymore but important is to let the notetaker join meetings without me being present Might be annoying for some people but so often I need only 10 min. of a meeting which happens on a driving distance. Sending a link in a chat that the other person opens is sooo handy!
- Collaborative / asynchronous notetaking - years ago I used a tool (can’t remember the name) which was asynchronous voice communication and it saved hours in some projects in which people were literally in all time-zones of the world. Back then I had to listen to or scroll through all transcripts and that’s the reason why the tool just didn’t make it. But now with LLM, this summary and hopefully soon audio clipping on request would solve so much of this (google are on the right track with notebooklm)
- Working with multiple recordings for generating insights - for example I have OPS meetings for each of the 6 companies once a week where we discuss problems and suggest solutions. It’s easy in some tools to just select the meetings and say - “put all problems and suggested solutions in a table for SOP discussion”
- Voice responding reader - driving or just tired of reading I use screen readers and it would be useful to have a voice reader for my notes
3 Likes
@Stefan_Stoyanov, many, many thanks indeed.
This is a well-researched and well-thought-out set of requirements.
I will find this very useful.
Well done.
respect
➤𝓜𝖆𝖝 𝖔’đť•đť–—đť–Žđť–Šđť–“
1 Like
Happy to finally give something back after all you’ve done for me/us
1 Like
@Max_OBrien
Here is what’s coming from Tana. Seems like a pretty good notetaker, aligned with some of my struggles:
Orchestration and syncing / birdging is very heavy and what i’m currently working in is figuring out how to take the “Actions” from a meeting and run that through a SOP (MD/Skill) for the role and the meeting - and then send; push these forward….e.g. some needs to go to Jira, some might require a few MS. Teams chats with some context….
Were currently on FireFlies….expensive, basic functionality…working on a coda pack where i download everything but now ideating and exploring on how to bring this step forward….and feels like it has to be an MCP with Claude/OpenAI etc. as i cant figure out how to use the SuperHuman AI……
1 Like
