Opportunities › Privacy-first local AI meeting transcription and notes › Proposed app
Roomkeep
Private meeting transcripts and notes that run on your own device
- Score
- 60
- confidence 61
- Evidence
- 17
- items on the page
- Rivals
- 8
- named in sources
- First version
- 45 d
- estimate
- Type
- Desktop and mobile
- PWA: on computers and phones alike
- Market since
- 2022
- Emerging market · AI estimate
ScreensConcept
Desktop and mobile: Meetings are recorded on laptops, while users review transcripts and notes on any device including phones between meetings. Sample screens on a computer and a phone, in the proposed brand. Their numbers are sample data.
Client call
Recording locally00:24:18
- Speaker 1
About this appAI proposal
Roomkeep records, transcribes and summarizes meetings inside your browser. Speech recognition and the summary model run on your own computer, so audio and text are not sent to a server. You get a transcript with speakers marked and a short set of notes, decisions and action items when the call ends.
It is made for people who hold sensitive conversations: lawyers, clinicians, finance and HR teams, and anyone whose company will not accept a cloud note taker. Install it as a web app on macOS or Windows, with nothing to host.
Unlike cloud assistants, notes are stored encrypted on the device. You can export to your own files or connect your own model key if you choose a larger model. Sharing is always a deliberate action.
- Tagline
- Meeting notes that never leave your device.
- Voice
- calmplaincareful
- Why this name
- A room you keep: the meeting stays inside it. Two plain words joined, easy to say, and it points at privacy without legal jargon.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Transcription on your device
Speech is turned into text by a model running in your browser. No audio is uploaded, and it keeps working offline once installed.
No member yet
Speaker labels
Each line is tagged to a speaker. Rename speakers once and the names carry through the transcript and notes.
No member yet
Summary, decisions, actions
A local language model writes a short summary, lists decisions and pulls out action items with owners.
No member yet
Encrypted local library
Meetings are stored encrypted on your device. Search across them, and delete one for good with a single tap.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#2F5D50Pine · primary#F4F1EAPaper · surface#1F2A28Ink · text#D9822BAmber · accent#3B6EA5Slate Blue · data
Typography
Aa
Roomkeep
ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
Meeting notes that never leave your device.
Fraunces for headlines and numbers, Figtree for everything else.
PricingAI proposal
Based on Fireflies' verified pricing page ($10 to $19 a month, plus free) and Otter's ($8.33 to $19.99, plus free). Roomkeep's tiers sit in that range and are a starting point to test.
Free
Free
- Small on-device model
- 5 meetings a month
- Export as text
Pro
$12 a month
- Larger on-device models
- Unlimited meetings
- Speaker labels and summaries
- Bring your own key
- No footer on exports
Team
$19 a month
- Everything in Pro
- Per seat pricing
- Shared templates
- Admin policy for what can be exported
What it could be worthEstimate
What a software business making this much sells for today, from real deals. Pick a monthly revenue.
If it makes
Worth about
$194,000
Sale range $132,000 to $263,000
$120,000 a year in revenue
From 87 software businesses listed and sold on Empire Flippers in the last year, revenue vetted by the broker: they sell for about 19× their monthly revenue, most between 13× and 26×.
Measured on profit, the middle deal is 32× monthly net profit: about 2.7 years of profit.
An estimate, not an offer or financial advice. A real price depends on profit, growth, churn, how much of the work is the owner's, and the buyer.
How it would growAI proposal
How the product would reach new customers through its own use, and how the first ones would be found. A plan to test, not a result.
- Built-in loop
- Shared outputs. Users export notes and action items as a file or link to attendees who do not use Roomkeep. Each export carries a small 'Notes kept on-device with Roomkeep' footer that can be switched off on paid plans.
- Do they become customers?
- Attendees who receive notes may be privacy-minded professionals who would pay, but a one-time reader may only see it. Test how many recipients click through and start a trial.
- First customers
- Solo and small-firm lawyers (1 to 10 people) who hold client calls and currently avoid cloud note takers for confidentiality reasons.
- Where to reach them
- Legal tech communities: Post in bar association forums and legal tech groups with a short demo showing the privacy receipt and offline use.
- Open-source local AI users: Share the approach in developer and self-hosting communities to gather feedback and early testers.
- Direct outreach: Email small firms with a one-page note on what leaves the device (nothing) and offer a 14-day trial.
- First milestone
- Ten small law firms pay for Pro after a trial, and at least half say privacy was the reason they chose it.
Why people already pay for this
- VerifiedGitHub repository gained 31537 stars in 651 days
- Public claimMeetily provides 4x faster transcription compared to existing solutions like Parakeet and Whisper.
- Public claimThe product emphasizes local processing and privacy as core features, requiring no cloud infrastructure.
- Public claimIt is positioned as the #1 self-hosted, open-source AI meeting note taker for macOS & Windows.
- Public claim
Tech stack
Every proposal builds on the same PWA base, installable from the browser on any phone or computer. Dashed boxes are what this app adds.
- AppNext.js and TypeScript
- PWAManifest and service worker
- UITailwind CSS
- Data and authSupabase Postgres
- PaymentsStripe
- EmailResend
- HostingVercel
- This app · On-device speech modelOpen speech model in WebGPU/WASMRuns speech recognition and speaker labeling in the browser via WebGPU or WebAssembly, so audio stays local.
- This app · On-device language modelSmall open LLM in WebGPUWrites summaries and action items locally, with no text sent to a server.
RisksAI inference
- TechnicalRunning speech and language models in a browser needs a capable computer and GPU support. Older machines may be slow, and live transcription speed is unproven.
- TechnicalCapturing both your microphone and the other participants' audio from a browser is limited by operating system and browser permissions, especially on macOS.
- CommercialFree open-source local meeting tools already exist and have attracted many users, so paying customers may need strong reasons such as polish, support and policies.
- AISmall on-device models may give weaker summaries and speaker labels than cloud services, which could hurt trust in the notes.
- RegulatoryRecording consent laws vary by place. Users need clear prompts, and no privacy claim should go beyond what the product actually does.
Everything marked AI proposal or AI inference was written by a model from the evidence on the opportunity page. The proposed name must be checked against trademark registers before it is used.