IntraGuard is an AI agent that monitors your financial data 24/7 — detecting duplicate payments, ghost employees, inflated invoices and supplier fraud before they become disasters.
No credit card required for trial · Xero direct connect · Your data never leaves your browser · BYOK pricing
No integrations required. Export a CSV from any system you already use and the agent does the rest.
Paste your Anthropic API key in Settings. You pay Anthropic directly for analysis — we charge only for platform access. No markup, no surprise bills.
Connect directly to Xero with one click and pull transactions and invoices automatically — or upload a CSV export from any accounting, payroll, or banking system.
The AI agent analyses your data using tool-based reasoning — flagging anomalies, scoring risk, and building evidence chains across 6 fraud domains.
Every finding comes with a forensic report, modus operandi analysis, and investigation steps. Export a professional PDF report. Critical findings trigger instant email alerts.
Works with your existing tools
IntraGuard monitors every major internal fraud vector simultaneously — not just the obvious ones.
Detects duplicate payments, round-number fraud, unusual timing patterns, and velocity anomalies in your bank and payment data.
Flags inflated invoices, single-supplier concentration, split purchasing under approval thresholds, and fictitious vendor indicators.
Cross-references employee and supplier data to surface name matches, shared addresses, and related-party transactions that indicate undisclosed conflicts.
Identifies ghost employees, payroll padding, suspicious bank account additions, and salary manipulation patterns in your payroll exports.
Monitors departmental spend against budget allocations, flags retrospective approvals, undocumented consulting fees, and systematic overspending.
Detects personal expenses on corporate accounts, asset misuse, weekend/holiday transaction clusters, and expense claims under approval thresholds.
IntraGuard uses a BYOK model — you bring your own Anthropic API key and pay them directly for analysis. We charge only for platform access.
+ your Anthropic API costs
+ your Anthropic API costs
Volume pricing available
A typical analysis scan on a 500-row CSV costs approximately $0.05–$0.15 in Anthropic API credits. Running daily scans on 4 data sources costs roughly $5–$15/month in API credits — far less than the platform fee. You get full transparency: check your Anthropic console for exact usage at any time.
Straight answers to the questions we hear most often.
Your data never leaves your browser — it is stored locally in your browser's localStorage and is never transmitted to IntraGuard's servers. The only external transmission happens when you run a scan: your CSV data is sent directly to Anthropic's API using your own API key. Anthropic's data handling is governed by their enterprise privacy policy. We have zero visibility into your data.
BYOK stands for Bring Your Own Key. Instead of IntraGuard paying for AI analysis and marking up the cost to you, you connect your own Anthropic API account. This means you pay Anthropic directly at their standard rates — with full transparency — and we charge a flat platform fee. No usage monitoring on our end, no surprise bills, no artificial caps on how much you can analyse.
IntraGuard accepts standard CSV files exported from any business system. This covers exports from Xero, QuickBooks, Sage, FreshBooks, MYOB, SAP, most banking portals, payroll systems (PaySpace, SimplePay, ADP), and any spreadsheet. If your system can export to CSV or Excel (saved as CSV), it works. We support 6 data types: bank statements, expense reports, payroll, invoices, purchase orders, and general ledger exports.
Accuracy depends on your detection sensitivity setting. At Medium sensitivity (recommended), the agent focuses on statistically significant anomalies with multiple corroborating signals — reducing false positives while catching genuine risks. At High sensitivity, it catches more patterns including lower-confidence signals, which will include more false positives. Every finding includes a risk score and confidence rating, and you can mark findings as False Positive with one click. The agent gets better context the more data you provide.
IntraGuard findings are an investigative starting point, not standalone legal evidence. The forensic reports identify patterns and generate investigation steps that a qualified forensic accountant or auditor should follow up on. The reports reference relevant legal and compliance frameworks (ACFE, IFRS, Companies Act) to guide formal investigation. For legal proceedings, findings should be validated by a certified forensic professional.
IntraGuard is designed for SMEs and mid-market businesses with 10–500 employees — the segment that suffers the highest fraud losses per incident but typically can't afford a dedicated internal audit team or forensic accountant. It's also used by accounting firms and forensic professionals as a rapid analysis tool for client work, and by CFOs and finance directors who want continuous monitoring without adding headcount.
No. If you can export a CSV from your bank or accounting software and paste an API key, you're set. The entire interface is built for finance professionals, not developers. The only slightly technical step is getting an Anthropic API key — which takes about 2 minutes at console.anthropic.com.
Yes — cancel anytime from your account settings or by emailing [email protected]. Your subscription remains active until the end of your current billing period. Since all your data is stored locally in your browser, there's nothing to "delete" on our end — your findings and sources exist only on your device. You can clear all local data at any time using the "Clear All Data" button in the app.
Upload your first CSV in under 5 minutes. The agent will show you what your business has been missing.
No credit card required · 7-day money-back guarantee · Cancel anytime