Oats is an AI meeting note-taking tool that is completely open, local, and free. Designed to not get in the way of your meetings, no bots, no subscription needed when running locally with on-device LLM. Available on both macOS and Windows. Enhanced transcription, multi-language support, speaker recognition, assessment, coaching, auto-tracking follow ups and more features via ariso.ai cloud backend.
@abod_rehman From our tests it's actually quite good, we have some novel techniques in here so it's not one-shotting the entire audio file which makes it easier for the local models to handle transcription.
@abod_rehman honestly cloud backend is a little better but the local model definitely do a fine job nowadays! You can try our cloud backend model as well. That's why give some flexibility for both
Report
Really curious about the speaker recognition part. does that work locally too or does it require the Ariso backend?
We built Oats because we believe people should have more control over their data. Privacy shouldn’t be something you trade away for convenience, and useful software shouldn’t require trusting a black box.
Oats is open source because we value FOSS: people should be able to inspect it, improve it, run it themselves, and keep ownership of their work. If you care about private, transparent tools, we’d love for you to check it out.
Report
okay, i really like this idea. not having a bot pop into every meeting is already a big win for me 😄
@colton_drake the local models do a decent job but honestly the cloud backend models is a little better but mostly for longer meetings, that's why we also provide a cloud backend ourselves, you can create a free account on ariso.ai to use the cloud backend.
Does Oats keep the original audio along with the transcript?
Oats
@avery_thompson2 yep, you can playback the original audio. It also supports voice profile to for speaker matching
Triforce Todos
if it is running fully local, how good is the transcription quality compared to something like otter that uses cloud?
Oats
@abod_rehman From our tests it's actually quite good, we have some novel techniques in here so it's not one-shotting the entire audio file which makes it easier for the local models to handle transcription.
Oats
@abod_rehman honestly cloud backend is a little better but the local model definitely do a fine job nowadays! You can try our cloud backend model as well. That's why give some flexibility for both
Really curious about the speaker recognition part. does that work locally too or does it require the Ariso backend?
Oats
@flora_kendall speaker recognition is currently handled by the Ariso backend :)
Oats
@flora_kendall We do speaker diarization locally and you can assign the names to their proper speaker identified.
No meeting bot is a really nice touch. does Oats work smoothly with Zoom, Meet and Teams?
Oats
@barnaby_lloyd It works with any audio being used on your laptop, it has no platform restrictions whatsovever because it runs entirely locally.
Oats
We built Oats because we believe people should have more control over their data. Privacy shouldn’t be something you trade away for convenience, and useful software shouldn’t require trusting a black box.
Oats is open source because we value FOSS: people should be able to inspect it, improve it, run it themselves, and keep ownership of their work. If you care about private, transparent tools, we’d love for you to check it out.
okay, i really like this idea. not having a bot pop into every meeting is already a big win for me 😄
Oats
@emily_carter18 Thanks! We realize it's sometimes tough and inconsistent for bots to join every call, hence why we built this!
Oats
@emily_carter18 definitely!
This solves one of my biggest issues with meeting tools. how accurate is the local transcription compared to cloud based tools?
Oats
@colton_drake the local models do a decent job but honestly the cloud backend models is a little better but mostly for longer meetings, that's why we also provide a cloud backend ourselves, you can create a free account on ariso.ai to use the cloud backend.