Verify the income.
Not just the pay stub.
ProofUp connects directly to your applicant's bank and payroll to confirm real earnings, then verifies their identity and catches forged documents. One of five verdicts in under 3 minutes, with the evidence behind it.
| Applicant | Unit | Rent | Income | Level | Method | Recommendation | Issue |
|---|---|---|---|---|---|---|---|
| M. Alvarez | 204 | $2,150 | $6,820 | 3.17x | Bank IncomeID + Selfie | Passes Requirements | |
| J. Patel | 118 | $1,895 | $3,910 | 2.06x | PaystubID + Selfie | Needs Review | Insufficient Info |
| D. Nguyen | 307 | $2,400 | $2,180 | 0.91x | Paystub | Fails Requirements | Fraud Detected |
| S. Williams | 412 | $1,750 | $5,640 | 3.22x | PayrollID + Selfie | Passes Requirements | |
| R. Okafor | 221 | $2,050 | $4,720 | 2.30x | Bank IncomeOffer Letter | Needs Review | Linking Error |
Trusted by property teams screening thousands of applicants
Show your work
Every flag has an explanation.
When an application doesn't pass, your team sees which document was flagged, why it failed, and which attempt triggered it. Every decision has a record.
- Full per-attempt verification history
- Specific reason for each document failure
- Verification method and source on every record
- Flag explanation in plain language
D. Nguyen
- Property
- Harmon Square
- Unit
- 307
- Monthly Rent
- $2,400.00
- Total Monthly Income
- $2,180.00
- Income Level
- 0.91x
- Report
- Download
- Applicant Name
- D. Nguyen
- Method Used
- Document Upload
- Verification Date
- 04/02/2026
- Documents Uploaded
- 3
- Invalid Docs Count
- 2
- Scanned Docs Count
- 1
The Problem
A forged pay stub looks exactly like a real one.
AI tools generate pay stubs with correct YTD math, real employer names, formatting indistinguishable from an actual ADP printout. Your team can't tell. Most screening software can't either.
- Pixel-perfect forgeries
- Modern PDF tools produce documents that pass every visual check. ProofUp reads the raw file data. The forgery shows up there.
- Identity mismatch
- A name and SSN on paper doesn't tell you who's actually at the showing. That's a separate check from the document.
- Unverified = risk
- Applicants approved without verified income default 4× more often. That number doesn't move until the screening changes.
ProofUp verifies the income before you hand over the keys.
Three Pillars
Income. Identity. Fraud.
Three things have to be true before you hand anyone a lease. Every application checks all three.
Income Verification
The applicant connects their bank and payroll via Plaid. Income comes from the account, not a document. For applicants who upload instead, AI reads deposit frequency and actual earnings. Works for gig workers and guarantors.
ID Verification
Government-issued ID photographed front and back. Live selfie taken. Facial recognition confirms the two match. The person at your door has to be the person on the application. Borrowed identities get caught here.
Fraud Detection
For applicants who upload, every PDF gets read at the data level. ProofUp checks the file structure: hidden text layers, metadata, generator fingerprints, post-generation edits. Tampering shows up in the data even when the document looks clean.
The Pipeline
Application to decision. Three minutes.
Every applicant runs the same checks. Your leasing team gets the result.
Applicant Gets the Link
Your team sends a link. The applicant opens it in any browser. No app needed. If they connect their accounts directly, no document uploads required.
Income Verified
Bank and payroll connected via Plaid, income pulled from the account. For applicants who upload instead, AI reads actual deposit history.
Identity & Fraud Checks
Government ID matched to a live selfie. Uploaded documents analyzed at the byte level for manipulation, generator fingerprints, and metadata mismatches.
Verdict Delivered
One of five verdicts, each with a risk level and the signals behind it. Your team makes the call with evidence in hand.
The verdict
A pass told your team nothing about the document behind it.
The old report returned one of three outcomes from the income math alone. A PDF pulled straight from ADP and a phone photo of a printout counted exactly the same. ProofUp now runs two more layers on top of that math: AI content analysis reads every uploaded document for tampering, and evidence grading records how the file reached you. You get one of five verdicts, in plain English, with the reason and the next step attached.
Jordan Lee and Taylor Morgan meet the income requirement, but the following items need a closer look before approving:
Jordan Lee, 2 paystubs: These were uploaded as a scan or phone photo, not downloaded from the payroll portal. Scans can hide edits a digital original would expose. Ask the applicant to re-upload an original PDF.
Verified income is 3.4x the $2,100.00/month rent, meeting your 3x minimum. However, at least one income document was scanned rather than downloaded from the source. ProofUp read it with OCR and counted the income, but a digital original would give you stronger evidence.
The report header as your team receives it. Applicant names, unit and figures are sample data.
The same income figure can arrive four ways. ProofUp counts all four and tells you which one you got.
Income comes from the account. There is no document in the path to forge.
The original file from the payroll portal. Metadata is intact, so edits are detectable.
A regenerated file. The original metadata is gone, and this type is easier to manipulate.
Read with OCR and counted. A scan can hide edits that a digital original would expose.
Every verdict carries a decision and a risk level.
Two of the five pass at medium risk. That distinction is the whole point: your leasing team can approve them, and your review team still knows which ones to sample. Select a row to see the action it implies.
Needs review is the verdict that costs you time. ProofUp will not clear an application carrying a fraud signal, a name mismatch or a broken bank link, so a person on your team has to open it and decide. That is the trade for never auto-approving a document nobody read.
Compare
Why teams switch to ProofUp.
| Manual Review | Legacy Screening | ProofUp | |
|---|---|---|---|
| Income Source | Applicant's word + a document | Document only | Direct from bank & payroll |
| Forged Document Detection | None | Limited checks | Byte-level + ML + rules |
| Identity Verification | Photocopy check | Basic KBA | Gov ID + live selfie match |
| Turnaround | Hours to days | Minutes | Under 3 minutes |
| Gig & Non-Traditional Income | No | No | Yes, bank-linked |
| Document Content Analysis | Human eyeball | None | AI reads every page for tampering |
| Evidence Quality Grading | None | None | Source-verified vs scan, per doc |
| Decision Output | Approve or decline | A score | 5 verdicts + risk level |
| Fraud Pattern Learning | Static | Static | Continuous, AI → rules |
Real Results
Results from portfolios that verify income at the source.
“We needed a solution that served a wide variety of applicants, including those without traditional forms of income. ProofUp's flexibility in supporting multiple verification methods, while maintaining rigorous fraud standards, has been a game changer for us. They have also been extremely receptive to our product feedback which has helped adoption grow rapidly throughout our portfolio.
Eddie Nientimp, Centralized Operations Manager, ResProp Management· Austin, TX · 24K+ units
Know the income is real.
Before the lease is signed.
Your team stops making calls based on a PDF. ProofUp pulls income from the account, matches identity to a live selfie, and reads every uploaded document at the data level. Three minutes per applicant.