Opportunities › AI context and memory for workplace automation › Proposed app
Wayfile
Private work memory your AI agents can read and you can review
- Score
- 21
- confidence 62
- Evidence
- 9
- items on the page
- Rivals
- 3
- named in sources
- First version
- 60 d
- estimate
- Type
- Desktop and mobile
- PWA: on computers and phones alike
- Market since
- 2020
- Established market · AI estimate
ScreensConcept
Desktop and mobile: Capture configuration and memory review happen at a desk, but the local capture helper monitors work across apps throughout the day including mobile context. Sample screens on a computer and a phone, in the proposed brand. Their numbers are sample data.
Memory
Capture running6.4 hrs
About this appAI proposal
Wayfile keeps a searchable memory of what you do across your apps, so AI agents get context without you pasting it in. A small local capture helper reads structured text from your windows instead of recording raw video, and stores it on your machine. The Wayfile app is where you review, trim and approve that memory.
You decide what an agent may see: by app, by project, by time window. Wayfile groups repeated steps into patterns and shows which ones look ready to automate, with the evidence behind each. Agents read the approved memory through one connection, with no set-up per tool.
It is for knowledge workers and small teams who want agents to act without constant micromanagement, and who worry about consent, local security, CPU load and noise. Those concerns are what Wayfile is built around.
- Tagline
- Your work, remembered locally. Shared only when you say.
- Voice
- CalmPlainCareful
- Why this name
- A way is the path you take through your day; a file is the record you keep of it. Two plain words, easy to say.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Structured capture, not raw video
The helper reads window text and app structure, skips duplicate frames and keeps the app context that plain video OCR loses.
No member yet
Approve what agents can see
Set rules by app, site, project or hours. Blocked sources are never indexed. Review and delete any slice of memory.
No member yet
Automation opportunities
Repeated sequences are grouped into patterns, with counts, time spent and the text snippets that support each one.
No member yet
One connection for agents
Agents query approved memory through one endpoint. No per-tool integration and no hand-picked sources to import.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#2F5D50Pine · primary#F6F1E7Paper · surface#1F2A24Ink · text#D9822BAmber · accent#4A6FA5Slate Blue · data
Typography
Aa
Wayfile
ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
Your work, remembered locally. Shared only when you say.
PricingAI proposal
The material gives no competitor prices. It says commercial use needs a licence, personal use is free, and enterprise adds services. The dollar amounts are a starting point to test.
Personal
Free
- Non-commercial, nonprofit, education, research
- Local memory and access rules
- One agent connection
Pro
$15 a month
- Commercial use licence
- Unlimited agent connections
- Opportunity reports
- Encrypted backup
Team
$25 a month
- Per seat, 5 seats minimum
- Shared opportunity report
- Admin consent policies
- Audit export
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 send opportunity reports, with evidence snippets they chose to include, to managers and teammates. Each report links to a page showing the pattern and the time it costs.
- Do they become customers?
- Recipients could become Team buyers, since they see the benefit, but a one-time read is only exposure. Test how many report viewers start a trial.
- First customers
- Operations leads and analysts at 10-50 person companies who do repetitive reporting in spreadsheets and email, and already use AI agents.
- Where to reach them
- Hacker News and developer forums: Show the privacy model and local-only design, then invite early users who raised consent and CPU concerns.
- Founder and ops communities: Offer a free two-week audit that turns one person's week into a ranked automation list.
- Direct outreach: Contact ops leads at small agencies and finance teams with a sample opportunity report.
- First milestone
- Ten teams complete a two-week trial and five pay for the commercial licence.
Why people already pay for this
- VerifiedHacker News post received 88 points and 67 comments
- AI inferenceAI agents need context about what people do on computers across applications to work autonomously without constant micromanagement.
- Public claimExisting tool-calling and MCP approaches for agent integration are insufficient: tool-calling is not autonomous enough and requires micromanagement, MCP is too static and difficult for non-technical users.
- Public claimPersistent knowledge systems like LLM-maintained wikis still require manual source selection and importing rather than autonomous capture of what people do.
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 · Local capture helperDesktop app using OS accessibility APIsA PWA cannot see other apps. A small helper captures structured text locally and pairs with the PWA.
- This app · Local indexOn-device SQLite with full-text and vector searchKeeps memory on the user's machine and fast to query without a server.
RisksAI inference
- PlatformA PWA cannot capture other apps. The product depends on a native helper per operating system, which needs OS permissions, signing and updates, and may be blocked by OS changes.
- RegulatoryRecording screens and audio raises consent and workplace monitoring rules that differ by region. Teams could misuse it for employee surveillance.
- TechnicalContinuous capture must stay light on CPU and battery and filter noise well, or users will switch it off. Past feedback named both.
- DataMemory may hold passwords, client data or health data. A leak or an agent reading too much would end trust, so redaction and local encryption must work from day one.
- CommercialExisting local-first screen memory tools already exist and may be free for personal use. Willingness to pay for commercial use is unproven.
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.