Earkeep
Maximise the value the team builds by keeping every ask that reaches it
Refinement, developer huddles, stakeholder demos: the backlog is decided out loud in rooms where you are facilitating, not typing, then delivered as acceptance criteria, an ordered backlog and a trail of why. Earkeep is an always-on transcriber that keeps every one of those rooms on your own device, with no bot present, so the ticket can match what was actually agreed.
No bot · No cloud · No record button
14-day full trial, no account required.
The conversations that decide it
Backlog refinement
What has to survive it is the acceptance criteria in the wording the Developers actually agreed to, including the edge case somebody raised at minute forty. You cannot facilitate a room, weigh a priority call and transcribe the agreement at the same time, so the version that reaches the ticket is the one you still remembered afterwards.
With Earkeep: connect Google Calendar, Apple Calendar, Outlook or any ICS URL and the ceremony starts, ends and titles its own span, so you spend it deciding, not writing. Afterwards the agreement is a transcript you can quote into the ticket instead of a paraphrase. How calendar spans work →
Developer huddle
What has to survive it is the exact clarification you gave, because the ticket has to match it and the code will be written against whichever version someone remembers. These happen several times a week, last five minutes, and are never scheduled, so nothing records them today.
With Earkeep: mic and system audio are already recording, all day, so a clarification given at somebody's desk lands exactly the way a calendared ceremony does. There is nothing to invite and no button to hit before the question gets asked. How always-on works →
Stakeholder demo
What has to survive it is the ask, with a date attached, and whether you agreed it becomes a backlog item. The half that matters most usually arrives after the demo ends, in a direct message or a follow-up huddle, from the one stakeholder who did not want to raise it in the room.
With Earkeep: that follow-up ask is on the same continuous timeline as the demo, and you select it as a span once you realise it mattered. Transcripts are kept as plain dated files in a folder you own, so "when did we agree to this?" is a search on your own disk two sprints later. From what was said to what you know →
The rest of the week
Those three are where the value is set. They are not where the week goes: for a month-long sprint the 2020 Scrum Guide alone timeboxes up to eight hours of planning, four of review and three of retrospective, and puts refinement near a tenth of the Developers' capacity on top. Flowtrace puts the average employee at 392 hours of meetings a year, about ten working weeks.
| Meeting | You | What has to survive it |
|---|---|---|
| Sprint planning | Run | The sprint goal, and the items actually forecast into the sprint backlog |
| Daily scrum | Attend | The blocker or scope question that needs a decision from you |
| Sprint review | Run | Stakeholder feedback, and the backlog change it causes |
| Sprint retrospective | Attend | Process changes affecting how you and the team work together |
| Roadmap or priority sync | Attend | Priority changes you have to reflect in the order of the backlog |
| 1:1 with the Scrum Master | Attend | Impediments that touch how work flows off the backlog |
| Ad-hoc or chat ask | Attend | Whether an informal request becomes a backlog item, and why |
Earkeep does not ask you to choose between them. Every recurring row above starts, ends and titles its own span from your calendar, so a full ceremony week ends as named transcripts with nothing left to file.
Writing them up is a separate job from keeping them, and the templates are free: templates for all five ceremonies, with the sprint goal recorded verbatim and refinement decisions attached to the items they concern, plus the planning template for the priority syncs above them.
See it in action
The ceremony that titles itself
Point Earkeep at the calendar your ceremonies already live on, and each one becomes a named span on its own. The huddle that was never scheduled is on the same timeline, unmarked until you decide it counted.
The criteria, drafted from what was agreed
Open a kept refinement, pick the agent you already use, and ask it for the acceptance criteria in the wording the room settled on. It reads the transcript, and you review a draft with the quotes still attached before any of it reaches the ticket.
Pricing: one payment, no subscription
€39
once, for one device, not a subscription.
No tiers, no per-seat billing, no recurring bill. Try it free for 14 days with full functionality, no email and no account required. See full pricing and policy details →
FAQ for product owners
Can Earkeep tell me exactly when a requirement was agreed to?
Yes. Every kept span is a dated, timestamped local transcript. If a stakeholder disputes a decision two sprints later, you or your agent can search the relevant window and produce the exact sentence and time instead of relying on memory.
Will it capture developer huddles that never get scheduled?
Yes. Because Earkeep records continuously in the background instead of only during scheduled events, an ad-hoc clarification at someone's desk or over a quick call is captured the same as a calendared ceremony; you mark the span afterward.
Does it work for in-person ceremonies, like refinement in a room with a laptop open?
Yes. Earkeep captures microphone audio as well as system audio, so in-person meetings are covered the same as video calls, as long as the laptop is present and Earkeep is running.
Do I need to change how I run ceremonies to use this?
No. Earkeep doesn't join the meeting or require anyone to open an app during it; it's already listening. You, or your calendar, mark which stretch of the day was the ceremony after it happens.
Does this replace Jira or Confluence?
No. Earkeep produces the transcript and, through in-app agents and MCP, the drafted comment, acceptance criteria, or decision summary. You still keep the backlog and decision docs in the tools your team already uses.
Is any of this sent to a vendor's servers?
No. Audio is processed only in memory and transcribed locally with Whisper or Parakeet; nothing is uploaded, and there's no account or cloud service in the loop, which also keeps IT and client stakeholders comfortable with the setup.
Stop losing what was agreed
14-day full trial, no account required.