1. Open the web app
Sign in to a paid Notes Recorder account. No desktop installer, extension, calendar connection, or meeting-bot invitation is required.
How Notes Recorder captures a meeting and turns it into follow-up.
Notes Recorder uses the microphone you select in the browser. It does not send a bot into the meeting and does not directly capture system or browser-tab audio. This guide explains the workflow plainly so you can decide whether it fits your setup.
Quick answer
The browser asks for microphone permission, records audio audible to that microphone, shows a live transcript when browser speech recognition is available, and uses the conversation to organize summaries, topic sections, decisions, risks, and action items. When you stop a supported recording with a live transcript, the app automatically saves the current meeting and attempts speaker separation by sending a recording of 25 MB or less through the private backend to OpenAI's audio-transcription API.
The five-step workflow
2. Grant microphone access
Select the device microphone that can hear the conversation. The browser and operating system remain in control of that permission.
3. Confirm permission to record
The app asks you to confirm that you are authorized to record and transcribe the meeting. Local laws and workplace rules still apply.
4. Record and review live
The US-English live transcript appears alongside a running summary and structured note when the browser supports Web Speech. Add rough notes without mixing them into the source transcript.
5. Finalize, save, and export
Stop the recording to auto-save the meeting and attempt speaker separation, then review the AI output, copy notes or transcript separately, or export the notes as Text, CSV, Word-compatible, or print-ready PDF output.
Human review remains required
AI notes can omit context or assign an owner or date incorrectly. Verify material details against the visible transcript before sharing or acting.
Capture compatibility matrix
This table describes the current product, not a future roadmap. Audio quality depends on the microphone, room, speaker volume, browser support, and network conditions.
| Meeting setup | Current fit | What to expect |
|---|---|---|
| In-person meeting around a laptop or phone | Strong fit | Choose a microphone that can hear everyone and place the device near the conversation. |
| Speakerphone call | Strong fit | The device microphone can hear both the room and the remote speaker if volume and placement are clear. |
| Zoom, Google Meet, or Teams played through speakers | Possible | The microphone must hear the computer speakers. Echo cancellation and volume can affect results. |
| Online call heard only through headphones | Not a fit | The microphone cannot reliably hear audio that remains inside the headphones. |
| Direct browser-tab or system-audio capture | Not supported | Notes Recorder currently uses microphone input and does not request tab or system audio. |
| Automatic calendar joining | Not supported | No bot joins on your behalf. You open the app and press Record yourself. |
| Mobile browser | Browser-dependent | Microphone recording and live speech recognition behavior vary by browser and device; test a short authorized sample first. |
| Language and browser | English-first | Live recognition is currently set to US English and final transcription defaults to English. Use a current Chrome or Safari release with microphone and Web Speech support; other browsers, accents, and languages are not guaranteed and should be tested first. |
| Legally authoritative or certified transcript | Not a fit | The transcript is AI-assisted product output and can contain errors; it is not a certified court or compliance record. |
What happens to the meeting data
- The browser receives microphone audio only after the user grants permission and starts recording. The browser's Web Speech implementation may use a browser or operating-system provider's network service for the live transcript.
- The browser uses live speech text to build the running transcript. Transcript text and meeting context are sent through the private backend to OpenAI's API for the running summary and draft note.
- When recording stops, the app automatically saves the meeting record. If there is a live transcript, AI is configured, and the audio upload is 25 MB or less, it also sends the audio through the backend to OpenAI's audio-transcription API to attempt speaker separation.
- The saved account-scoped record contains the transcript, structured notes, rough notes, and meeting metadata. Notes Recorder does not store the audio file in saved meeting history; a current-session audio download is available in the browser.
- Customers can delete saved meetings, copy notes and transcript separately, and export the notes in supported formats.
Read the security overview and privacy policy for the current service commitments. Do not record confidential or regulated information unless your organization has approved the workflow.
What the AI note includes
Meeting title and overview
A title, high-level summary, decisions, and risks based on the recorded conversation.
Topic sections
Discussion grouped into readable sections rather than one undifferentiated transcript.
Action items
The AI is instructed to attach owners and due dates only when names and dates are supported by the transcript. Ambiguous fields may remain unspecified, and human review is required.
Source transcript
The source stays separate so you can review context, correct speaker labels, and refresh the note after an edit.
When to choose another tool
Choose a system-audio or desktop recorder if headphones are mandatory. Choose a bot-based platform if automatic meeting joining is the priority. Choose an enterprise platform if you need SSO, administrative retention policies, legal hold, CRM write-back, or a signed compliance agreement that Notes Recorder does not currently provide.
Ready to evaluate the fit?
Review the current plans, compare the workflow using the buyer checklist, or open the app when your meeting setup matches the capture method above.