Opportunities › AI-native knowledge base and note-taking › Proposed app
Larder
Notes your AI assistants can search and add to, self-hosted or hosted
- Score
- 47
- confidence 50
- Evidence
- 4
- items on the page
- Rivals
- 2
- named in sources
- First version
- 45 d
- estimate
- Type
- Desktop and mobile
- PWA: on computers and phones alike
- Market since
- 2004
- Mature market · AI estimate
ScreensConcept
Desktop and mobile: Knowledge workers reference notes while coding at a desk but also capture and search them on mobile with AI assistants. Sample screens on a computer and a phone, in the proposed brand. Their numbers are sample data.
Notebook
Offline, will sync128 notes
- Trip planning 2027
About this appAI proposal
Larder is a notebook for people who work with AI assistants. Notes are plain Markdown, and any assistant that speaks MCP (Model Context Protocol) can search them, read them and add new ones, with permissions you set per notebook.
It works offline as a PWA and syncs when you reconnect. Run it on your own Docker host or Cloudflare account, or use the hosted version if you would rather not manage servers.
It is for individuals and small teams who keep a personal knowledge base and want AI in it without a closed tool or lock-in. Export everything as Markdown at any time.
- Tagline
- A well-stocked shelf of notes, open to your AI and closed to lock-in.
- Voice
- CalmPlainPractical
- Why this name
- A larder is a store you keep for later and draw from when needed, like a knowledge base. Short, pronounceable, and not tied to any note app.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Built-in MCP server
Connect an AI assistant with one URL and a token. It can search, read and write notes in the notebooks you allow.
No member yet
Per-notebook access
Set each notebook to read-only, write-allowed or hidden for every connected assistant. Revoke a token any time.
No member yet
Offline-first
Read and write with no connection. Changes merge when your device reconnects.
No member yet
Self-host or hosted
Deploy with Docker or to your own Cloudflare account, or let us run it. Same app either way.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#2F5D50Pine · primary#F6F1E7Flour · surface#2B2A26Charcoal · text#D9822BApricot · accent#5B7FA3Slate blue · data
Typography
Aa
Larder
ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
A well-stocked shelf of notes, open to your AI and closed to lock-in.
PricingAI proposal
The material gives no competitor prices. These tiers are a starting point to test, built on the idea that self-hosting removes software cost.
Self-hosted
Free
- Run it yourself with Docker or Cloudflare
- MCP server and offline PWA
- Markdown export
- Community support
Hosted
$6 a month
- We run and back it up
- MCP server for your assistants
- Sync across devices
- Agent activity log
Team
$5 a month
- Price per person, hosted
- Shared notebooks
- Per-notebook assistant access
- Priority support
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
- Users invite others. Team members are invited into shared notebooks, and each invitee also gets their own notebooks and assistant connections.
- Do they become customers?
- Invitees can become paying users since they use the full product. Test whether invited teammates start their own hosted accounts or stay as free guests.
- First customers
- Solo developers and technical writers who already use AI assistants daily and keep notes in Markdown or Evernote.
- Where to reach them
- GitHub and open-source communities: Publish the self-host template and a short guide to connecting an assistant over MCP.
- Developer forums and newsletters: Post a walkthrough of moving Evernote notes into an AI-connected notebook.
- First milestone
- Fifty self-hosted installs and ten paying hosted users within three months of launch.
Why people already pay for this
- VerifiedGitHub repository gained 2179 stars in 106 days
- Public claimUsers seek note-taking and knowledge management solutions that integrate with AI agents
- Public claimCost and infrastructure flexibility matter to users in this market
- AI inferenceThere is demand for self-hosted and offline-first knowledge management
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 · Offline syncLocal-first storage with merge sync (CRDT or similar)Notes must be editable offline and merge without losing edits.
- This app · MCP serverMCP endpoint with scoped tokensLets AI assistants search and write notes under per-notebook permissions.
RisksAI inference
- TechnicalOffline sync with merge across devices and assistant edits is hard to get right; lost or conflicting edits would destroy trust quickly.
- CommercialSelf-hosting is free by design, so few users may pay for hosting. Willingness to pay needs testing.
- PlatformMCP is a young protocol and clients may change how they connect; the server will need ongoing updates.
- AIAssistants with write access could overwrite or leak notes. Scoped tokens, logs and undo must be solid.
- DataEvernote import depends on export formats, and imperfect imports could lose tags, attachments or dates.
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.