How AuditPulsar works

Not a rules engine. Not a checklist tool. A detection system trained on real general ledger data that tells you which transactions to investigate first.

From raw ledger data to prioritized findings

Four stages. Runs in under 90 seconds for engagements up to 500,000 journal entries.

01

Data Ingestion

Connect directly to SAP S/4HANA via BAPI RFC, Oracle NetSuite via SuiteQL, or QuickBooks Enterprise via the Intuit Data API. For other ERPs, the CSV ingestion pipeline normalizes account codes, posting dates, and cost centers before processing begins.

02

Statistical Baseline

AuditPulsar builds a transaction baseline from 12 months of history. It applies Benford's Law distribution analysis to detect digit-frequency anomalies, and calculates posting-time distributions to identify after-hours entries that fall outside normal approval workflows.

03

ML Anomaly Scoring

A gradient-boosted classifier trained on 14.7M labeled journal entries scores each transaction on a 0–100 scale. Features include account-to-account relationship patterns, preparer frequency, reversing entry cadence, and period-end concentration. High-scoring items surface in the findings queue first.

04

Workpaper Export

Each finding package includes the source journal entry, account code, preparer ID, approval chain, and auditor notes. Export as PDF workpaper, Excel, or push directly to cloud document management systems via REST API. Every export is timestamped and signed.

What the platform can do

Designed around the actual workflow of an audit engagement — not a general analytics tool adapted for audit.

Multi-Entity Support

Run anomaly scans across consolidated ledgers for clients with multiple subsidiaries. AuditPulsar maintains separate baselines per legal entity and can surface intercompany elimination discrepancies that manually prepared consolidations often obscure.

Supports both elimination method and proportional consolidation. Identifies round-trip transactions that inflate revenue across entities — a common area of management override risk under AS 2201.

Period-End Analysis

The engine applies heightened scrutiny to journal entries posted in the final 3 business days of each reporting period. These entries represent a disproportionate share of audit findings — accrual reversals, inventory adjustments, and manual revenue entries that management posts to hit targets.

Produces a period-end concentration report segmented by account type (revenue, liability, asset) and preparer. Reviewers can spot patterns across multiple periods in a single view.

Trial Balance Reconciliation

Traces every trial balance account balance back to its component journal entries. Cross-checks subsidiary ledger totals (accounts receivable subledger, fixed asset register) against the general ledger control accounts, flagging any unexplained differences above the materiality threshold you configure.

Reconciliation reports reference the specific source transactions responsible for each difference. No more exporting to Excel to trace a $12,000 AR subledger gap.

Preparer Behavior Analysis

Builds a posting profile for each user account in the client's ERP. Flags entries where the preparer has unusually high authority (full journal access with no required approval), entries posted under accounts the preparer rarely touches, and entries logged outside normal business hours.

User behavior analytics are aggregated by role and department, giving auditors a defensible basis for extending or restricting sample sizes by department.

SOC 2 Type II Compliance

Client financial data is processed in isolated compute environments. No data persists after the scan completes. AuditPulsar operates under a shared responsibility model — your firm controls the connection credentials and the output documents. Our infrastructure and security controls are audited annually by a Big-4 assurance team.

BAA available for firms requiring additional contractual data protections beyond SOC 2.

Continuous Monitoring Mode

For clients under ongoing internal audit or quarterly review engagements, AuditPulsar offers a continuous monitoring subscription. Ledger data syncs weekly. New anomalies trigger email alerts to the engagement team before month-end close, reducing the volume of adjusting entries that accumulate until year-end.

Alert thresholds are configurable per account class. Revenue and cash accounts can be monitored daily; fixed assets monthly.

Works with the ERPs your clients already use

Certified connectors for the most common systems. Custom integrations via REST API for everything else.

SAP S/4HANA

Connects via RFC/BAPI. Pulls FI document tables (BKPF, BSEG) directly. No middleware or SAP Basis team involvement required after initial OAuth setup.

Oracle NetSuite

Uses SuiteQL REST endpoint. Fetches transaction records, subsidiary structure, and account metadata in a single authenticated session. Respects NetSuite role-based permissions.

QuickBooks & Xero

Supports QuickBooks Enterprise and Xero via their respective accounting APIs. Suitable for smaller clients where full ERP connectivity isn't available.

Also supports: Microsoft Dynamics 365, Sage Intacct, Epicor ERP, and flat CSV/Excel import from any system.

Built for the scale of real audit engagements

500K
Max journal entries per scan (Growth plan)
2.4s
Average scan time per 10,000 entries
99.97%
Platform uptime (rolling 12 months)
AES-256
Encryption at rest and in transit

Ready to run your first scan?

Schedule a 30-minute demo with a member of the AuditPulsar team. We'll walk through the platform using a sample ledger matching your client size.

Request a Demo