Opportunities › Tax document classification and processing automation › Proposed app
Pigeonhole
Sort scanned tax documents by IRS form type and review only the uncertain pages.
- Score
- 43
- confidence 43
- Evidence
- 4
- items on the page
- Rivals
- 0
- named in sources
- First version
- 35 d
- estimate
- Type
- Desktop app
- PWA: runs in the browser, installs on Mac and Windows
- Market since
- 2010
- Mature market · AI estimate
ScreensConcept
Desktop app: Tax professionals process batches of documents, review classifications, and audit missing forms across multiple pages. Sample screens on a computer, in the proposed brand. Their numbers are sample data.
Smith household
138 of 142 sorted142 pages
About this appAI proposal
Pigeonhole reads each page of an incoming tax document batch and labels it with its IRS form type. Drop in a scanned client folder or send pages through the API, and get back a sorted, indexed set ready for preparation or filing.
It is built for tax professionals, accounting firms and tax software teams who handle large volumes of mixed documents. Pages the model is unsure about go to a short review queue instead of slipping through, so people only look at the exceptions.
Public claims on GitHub describe a classifier reaching 100% strict accuracy across 261 IRS forms at about $0.001 per page. These are unverified; the first version should measure accuracy on your own documents before you rely on it.
- Tagline
- Every tax page, sorted to the right form before anyone opens it.
- Voice
- PlainPreciseCalm
- Why this name
- Like sorting mail into pigeonholes: a sorter that puts each page in its slot. Short, pronounceable, and not tied to any tax product name.
In the first version
Members say which features they would use. It tells PWAS what people want; it changes no score.
Form-level page classification
Each page is labeled with its IRS form type, including pages in the middle of a mixed, unsorted scan.
No member yet
Batch intake
Upload a folder or PDF of any size, or forward by email. Results arrive as a sorted, indexed set per client.
No member yet
Confidence and review queue
Low-confidence pages are held for a quick human check, with the model's top guesses shown side by side.
No member yet
Missing document checklist
Compare what arrived to what a client's return needs and see which forms are still missing.
No member yet
Brand identityAI proposal
Check the name against trademark registers before using it.
Palette
#1F4D3ALedger Green · primary#F7F3EAPaper · surface#1B1F1DInk · text#D9622BStamp Orange · accent#8FAE9BSage · data
Typography
Aa
Pigeonhole
ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
Every tax page, sorted to the right form before anyone opens it.
PricingAI proposal
The material gives no competitor prices. Tiers are a starting point to test, informed only by the public claim of roughly $0.001 per page operating cost, which is unverified.
Trial
Free
- 200 pages a month
- Review queue
- Accuracy report on one sample
Firm
$49 a month
- 10,000 pages a month
- Email intake and checklists
- Review queue
- Extra pages billed per page
Volume
$299 a month
- 150,000 pages a month
- API and webhooks
- Priority support
Partner
Contact us
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
- Partners and integrations. Tax software and practice-management tools embed the classifier through an API and white-label option, so intake in their products runs on it. Firms also connect it to shared inboxes and cloud folders.
- Do they become customers?
- Integration partners bring end users who see it only as a feature. Test whether partner firms will pay directly for the standalone dashboard, or only through the partner's contract.
- First customers
- Small and mid-size US accounting firms (5 to 50 staff) that handle individual returns in busy season and sort client document uploads by hand.
- Where to reach them
- Accounting communities: Offer free accuracy checks on a firm's own sample batch in tax-preparer forums and groups.
- Tax software vendors: Pitch the API to small tax-prep and document-portal products that lack their own classifier.
- Open-source community: Engage the GitHub audience around the public classifier and invite them to a hosted beta.
- First milestone
- Five firms each run a real batch of 1,000+ pages, confirm accuracy against their own labels, and three agree to pay.
Why people already pay for this
- VerifiedGitHub repository gained 505 stars in 25 days
- Public claimThe classifier achieves 100% strict accuracy across 261 IRS forms, indicating high reliability for a critical compliance task.
- Public claimOperating cost of approximately $0.001 per page processed suggests a viable, low-cost service model.
- AI inferenceThe project has gained attention with 505 stars and 64 forks on GitHub, indicating user interest in tax document classification solutions.
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 · Document classification modelHosted page-image classifier with GPU or API inferencePage-level form recognition needs model inference beyond the shared stack.
- This app · File storageEncrypted object storage with expiry rulesScanned tax files need encryption at rest and automatic deletion.
RisksAI inference
- AIThe 100% strict accuracy figure is a public claim. Real scans (skewed, faxed, handwritten) may score lower, and a misfiled form in a return carries real consequences.
- RegulatoryTax documents hold SSNs and income data. Firms will expect encryption, retention limits and written security practices, and may face IRS data-security rules.
- CommercialAt about $0.001 per page claimed cost, a rival can undercut easily. Value must come from workflow and integrations rather than the model alone.
- TechnicalNew and revised IRS forms appear each year, so the model needs a retraining and testing routine before every filing season.
- Intellectual propertyThe model may build on the public GitHub project, so its license and training data terms must be checked before commercial use.
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.