AbaAiCheck
Guide Agency Owners, Compliance Officers & Auditors

Bulk Audit Forensics & FWA Prevention Guide

How to run a batch forensic analysis of session notes to detect cloned content, overlapping sessions and compliance risks before payers do.

In-memory processingBAA availableTLS encryptionHuman review required

Security & data handling

In-Memory Processing, Nothing Stored

Uploaded files are read in memory to build the report and are then discarded: the notes and the report are not saved to our database or file storage. The notes selected for Stage 2 are sent to our HIPAA-eligible AI service only for that analysis, under our Business Associate Agreement.

Pre-Audit Aid

Findings are an internal risk-mitigation tool for human compliance reviewers. They are not a legal determination of fraud, waste or abuse, and every finding should be verified against the source documentation.

The report lives only in your browser tab: it is not available after you leave or reload the page, so act on it (or print the page) before closing. See HIPAA & Compliance for our safeguards.

Step-by-Step Workflow

  1. 1

    Step 1

    Prepare & Upload the Batch

    • Export the session notes you want to review from your EHR (for example, a month of notes for one provider or one payer).
    • Supported formats: ZIP, PDF, DOCX and TXT. A ZIP can hold any mix of PDF, DOCX and TXT files.
    • Capacity per batch: up to 20 files and 200 notes, up to 25 MB per file (a ZIP can expand to 100 MB).
    • In the web app, open Billing & Revenue → Bulk Audit Forensics, then drag and drop the batch into the upload zone or click to select the files.

    Available to agency owners and Billing Specialists. Use labeled fields in your export (date of service, provider, time in / time out, units) so every check can run.

  2. 2

    Step 2

    Stage 1: Local Pre-Filter

    • The engine reads every note and runs rule-based checks across the whole batch in a few seconds:
    • In-batch clone detection: how similar each narrative is to every other note in the batch.
    • Overlapping sessions: the same provider billed for overlapping times, and more than 12 hours billed by one provider on one day.
    • Units versus time: units billed above what the session time supports (15 minutes = 1 unit).
    • Signatures: notes that appear unsigned or were signed late (more than 24 hours after the date of service by default; your agency can change this rule).
    • Missing data: notes without a labeled date of service, provider or session time, which cannot be verified.
  3. 3

    Step 3

    Stage 2: AI Batch Forensic Review

    • Notes flagged as high or medium risk by the pre-filter go to the Bulk Audit Engine (AI), highest risk first: up to 20 notes per batch and 100 notes per user per day.
    • The AI reviews cloned or templated narratives, impossible time overlaps, late signatures and missing dates, and unit/time math, and explains each finding note by note.
    • If the AI is unavailable or the daily limit is reached, you still get the complete pre-filter report.

    The AI only works with the data in the batch: it does not invent dates, names or numbers, and it does not repeat client names in its findings.

  4. 4

    Step 4

    Interpret the Batch Forensics Summary

    • The dashboard opens with four forensic metrics (detailed below), followed by the note-by-note report with each note’s risk level, flags and AI findings.
    • Start with the high-risk notes, confirm each finding against the source documentation, and correct or withdraw notes before they are billed or requested in a payer audit.

The Four Forensic Metrics

What the Batch Forensics Summary at the top of the report means.

#

1. Total Notes Analyzed

Distinct session notes read from the batch (files that could not be read are listed separately).

%

2. Cloned Rate

Share of notes whose narrative is at least 80% similar to another note in the same batch.

!

3. High Risk Count

Notes with a critical flag: a clone of 90% or more, overlapping sessions, more than 12 billed hours in a day, or units above what the time supports.

$

4. Potential Exposure

Estimated amount at risk: high-risk notes × an estimated amount per note set by your administrator. An estimate to prioritize review, not a calculated loss.

Risk Levels

How the pre-filter assigns each note’s risk before the AI review.

Risk levelTriggered by
HighClone ≥ 90%, overlapping sessions, > 12 billed hours per provider per day, units above session time
MediumClone 80–89%, late signature, note appears unsigned
LowNo critical flags (missing-data notices may still appear)

Run your first forensic batch

Bulk Audit Forensics is in the Billing & Revenue menu of the AbaAiCheck web app.

Bulk Audit Forensics supports internal pre-audit review. It does not replace your compliance program, legal counsel or payer-specific guidance, does not determine fraud, waste or abuse, and does not guarantee audit outcomes. Exposure figures are estimates. Your agency remains responsible for every note it bills.