Feature Deep Dive

Your early-warning system for financial surprises.

Anomaly Radar uses advanced machine learning to monitor your cash flow in real-time, flagging suspicious transactions, duplicate payments, and category drift before they impact your bottom line.

Enable Anomaly Radar View Documentation
JDALMR
Active on 2,400+ accounts
Voxel Anomaly Radar interface highlighting a detected duplicate transaction
High Risk Detected $4,200 duplicate
Detection Capabilities

What the Radar finds.

Duplicate Payments

Identifies identical or near-identical transaction amounts and descriptions occurring within the same time window across connected accounts.

Unusual Vendors

Finds charges from merchants you've never interacted with before, or sudden spikes in spending with long-standing vendors.

Spike Patterns

Monitors non-linear spending velocity. A sudden 300% increase in operational costs triggers an immediate alert.

Category Drift

Detects when a transaction suddenly shifts categories (e.g., a "Software" charge appearing under "Travel"), which can indicate budget leakage.

Under the Hood

How the AI model works.

Anomaly Radar doesn't just look at yesterday's balance; it builds a dynamic model of your unique financial behavior.

1. Baseline Learning Period

During the first 30 days, Voxel analyzes your historical data to establish a "normal" spending range for every category and vendor. This learns your seasonal variations and routine recurring payments.

2. Sensitivity Scoring

The model assigns a confidence score (0-100%) to every transaction. A $100 expense might be flagged if it's 3x the usual amount, but a $10,000 expense would require only a 1.1x variance to trigger an alert.

3. Confidence Thresholds

You control the sensitivity. High confidence alerts (99%+) are rare and specific. Low confidence alerts (60%+) catch broader trends and give you context before they become critical issues.

Setup Guide

Setting up your first alert rules.

Step 1: Define Your Baseline

Navigate to Settings > Anomaly Radar. Voxel will automatically request read-only access to your connected accounts to analyze the last 90 days of transaction history.

> Initiating baseline analysis...
> 1,243 transactions scanned.
> Baseline established for "Marketing Spend".

Step 2: Configure Thresholds

Use the sliders to set acceptable variances. For example, allow a 20% variance for "Office Supplies" but only 5% for "Software Subscriptions".

Category Variance Limit 15%
Alert Channels

Stay informed instantly.

📧

Email Digest

Daily or weekly summaries of high-priority anomalies. Perfect for closing the books.

💬

Slack Integration

Direct channels for critical alerts. Tag your finance team instantly.

📱

Push Notifications

Instant alerts to your mobile device for urgent security breaches.

🔗

Webhooks

Send anomaly data to your own CRM or internal ticketing system.

Case Study

Catching the $4,200 duplicate.

The Situation

TechFlow Inc. had an automated SaaS renewal scheduled for $4,200. Due to a billing configuration error, the charge was processed twice within 24 hours. Voxel's anomaly radar caught the second charge immediately.

Impact: $4,200 recovered.
Time to detection: < 4 minutes.

The Resolution

The automated alert was routed to the CEO's Slack channel. The finance team verified the transaction, initiated a dispute with the vendor, and received the funds back within 5 business days. Without Voxel, this would have gone unnoticed until the next month's reconciliation.

Fine-Tuning

Tuning false positives.

Anomaly Radar learns from you. If a legitimate transaction triggers a warning, simply mark it as "Not an Anomaly" to refine the model's accuracy for your specific organization.

Feedback Loop Active Model Re-trained Daily Smart Exclusions
FAQ

Common questions.

For most accounts, the baseline is established within the first 14 days of connection. However, for accounts with highly volatile spending (e.g., seasonal retail), we recommend waiting 30 days before adjusting sensitivity thresholds for maximum accuracy.
Yes. On the Pro plan and above, you can create custom rules, such as "Alert if any transaction exceeds $10,000" or "Alert if more than 3 transactions occur on a Sunday."
Anomaly Radar is a proactive monitoring tool that catches financial inconsistencies. For high-value enterprise accounts, we recommend pairing it with our dedicated Fraud Shield module for real-time transaction blocking.
Ready to secure your finances?

Enable Anomaly Radar on your account today.

Stop guessing and start seeing. Get 14 days of full access to all intelligence features.

Enable Radar

No credit card required · Enterprise-grade security