Is PaperAI HIPAA compliant?

No. PaperAI is not HIPAA-compliant and does not sign Business Associate Agreements. Do not upload documents containing PHI. PaperAI is suitable for de-identified workflows only — training data preparation, code-library research, and denial-pattern analysis on documents with all 18 HIPAA identifiers removed. If you need to process PHI, use a vendor that holds a signed BAA today.

How does PaperAI handle different payer EOB formats?

Every payer (BCBS, UHC, Aetna, CMS, Humana) has subtle EOB variations. PaperAI's vision AI reads them semantically rather than via fixed templates, so the common fields extract reliably out of the box. We add payer-specific sub-templates as customer needs evolve.

Does PaperAI provide clinical or coding decisions?

No. PaperAI extracts data only — we never recommend clinical or coding decisions. All extracted codes, amounts, and denial reasons should be reviewed by qualified billing or clinical staff before submission or appeal.

Can it flag denial reason codes (CARC/RARC) on EOBs?

Yes. PaperAI extracts CARC (Claim Adjustment Reason Codes) and RARC (Remittance Advice Remark Codes) from EOBs as structured typed fields. Filter denials by code as a batch — 'show me every CO-45 denial in the last 30 days' — so your appeals team works dollars instead of reading EOBs.

How does it handle handwritten edits on claims?

Premium AI models detect handwritten edits, scratched-out values, and pen-and-ink corrections. These get flagged as low confidence in the side-by-side view so your billing reviewer verifies the corrected value before the claim posts to your PM system.

Can PaperAI extract data from secondary payer EOBs?

Yes. Secondary payer EOBs that arrive as faxes (the documents your clearinghouse usually doesn't process) extract the same way as primary EOBs. This is often the highest-ROI use case — secondary payer denials are where manual review misses the most dollars.

What if a payer changes their EOB format?

Because PaperAI reads layouts semantically rather than via templates, payer format changes typically don't break extraction. If a specific field starts extracting with low confidence after a format change, our team adds a payer-specific sub-template within a few business days.

How does PaperAI compare to Waystar, Availity, or Inovalon?

Those are full RCM clearinghouses with direct payer connectivity, eligibility verification, and end-to-end claim lifecycle. PaperAI is the extraction step for inbound documents the clearinghouse doesn't process — faxed EOBs, paper claims, downloaded payer remittances. Pairs with your existing RCM stack rather than replacing it. See /compare/medical-claims-processing-tools.

Can I run this for a multi-tenant billing service?

Yes. Multi-tenant data isolation by org means each practice client's claims stay separated. Role-based access keeps your team's view scoped per client. Business and Scale plans support the per-client extraction volumes typical of billing services.

How does pricing work?

Each claim or EOB page is approximately 2 credits on standard models, 4 credits on premium. The $19/mo Pro plan includes 1,000 credits (~500 claim pages). Free tier includes 100 credits one-time to validate accuracy on real EOBs before paying anything.

PaperAI is not HIPAA-compliant and does not sign Business Associate Agreements.

Do not upload PHI. This vertical is intended for de-identified workflows only — training data prep, denial-code library research, coder education, and pattern analysis on scrubbed sample EOBs. If your workflow requires PHI handling, choose a vendor that holds a signed BAA today. See /trust for current status.

Claim data extraction (de-identified)

Find every appealable denial in a stack of de-identified EOBs — in 30 seconds.

For RCM teams running training, coder onboarding, and denial-pattern research on scrubbed sample EOBs and CMS-1500s. Surfaces CARC/RARC codes, billed-vs-paid, and CPT/ICD lines as structured fields. PaperAI is not HIPAA-compliant — see hero banner.

Try on a de-identified claim

PaperAI is not HIPAA-compliant and does not sign BAAs. Do not upload PHI. Use de-identified claims, sample EOBs, or training data only.

What teams actually use this for

Healthcare RCM teams spend hours every day re-keying CMS-1500, UB-04, and EOB documents — most still arriving as faxes, scans, or downloaded PDFs from payer portals. Every payer (BCBS, UHC, Aetna, CMS) has subtle EOB variations, and a single transposition error on a CPT code or paid amount can mean a missed appeal window.

PaperAI extracts the structured data your billing system needs: patient demographics, dates of service, CPT and ICD codes, billed and paid amounts, denial reason codes (CARC/RARC), and adjustment details. Confidence scoring ensures uncertain fields get flagged for clinical or billing review before they hit your practice management system.

**PaperAI is not HIPAA-compliant and does not sign Business Associate Agreements.** Today the product is suitable only for de-identified workflows — training data preparation for billing teams, CARC/RARC code library research, denial-pattern analysis on scrubbed sample EOBs, and coder onboarding. Do not upload documents containing PHI (patient names, MRNs, DOBs, addresses, or any of the 18 HIPAA identifiers). If your workflow requires PHI handling, choose a vendor that holds a signed BAA today.

Sample input → output

A de-identified EOB in. Structured denial codes out.

Sample de-identified EOB (training / illustration)

SAMPLE PAYER — DEMONSTRATION ONLY
Explanation of Benefits — Provider Copy
Provider: [DE-IDENTIFIED]              NPI: XXXXXXXXXX
Patient: [TEST RECORD #00001]

Svc Date    CPT     ICD-10   Billed   Allowed   Paid     CARC
03/18/2026  99214   M54.5    $285.00  $142.50   $142.50  --
03/18/2026  73564   M54.5    $180.00  $0.00     $0.00    CO-45

CARC CO-45 = Charge exceeds fee schedule / max allowable
Total Billed: $465.00    Total Paid: $142.50

Structured claim record with denial codes

json

{
  "payer": "[Sample Payer]",
  "provider": "[de-identified]",
  "provider_npi": "XXXXXXXXXX",
  "patient_record_id": "TEST-00001",
  "claim_lines": [
    { "dos": "2026-03-18", "cpt": "99214", "icd": "M54.5",
      "billed": 285.00, "allowed": 142.50, "paid": 142.50, "carc": null },
    { "dos": "2026-03-18", "cpt": "73564", "icd": "M54.5",
      "billed": 180.00, "allowed": 0.00, "paid": 0.00, "carc": "CO-45" }
  ],
  "total_billed": 465.00, "total_paid": 142.50
}

Common questions

Answers focused on conversion quality, team workflows, and roadmap clarity.

PaperAI processes CMS-1500 (professional claims), UB-04 (institutional claims), and payer EOBs/ERAs. The template is built on the common 1500/UB-04/EOB structure; payer-specific sub-templates can be added as your team needs them.

Validate the workflow on de-identified claims first.

PaperAI is not HIPAA-compliant. Use 100 free credits on scrubbed sample EOBs, training-data prep, or code-library research. Do not upload PHI.

Try on de-identified claims