Alerts triaged by real machine learning.
ICAT risk-scores 327 employees and 580 vendors across Travel & Expense and Procure-to-Pay, maintains a registry of 19 trained model versions, and explicitly measures false-positive reduction by model.
Smart Alert Triage (ICAT) is Moderor's ML-driven alert triage product for financial operations. It scores employees and vendors on continuous risk spectrums, flags outlier transactions, and maintains a versioned ML model registry — measuring false-positive reduction per model version.
- 327 EMPLOYEES RISK-SCORED
- 580 VENDORS ANALYSED
- 19 TRAINED MODEL VERSIONS
- FP REDUCTION MEASURED PER MODEL
Capabilities
/ FEATURESTravel & Expense triage
327 employees scored on a continuous risk spectrum (15 high / 89 medium / 223 low) with transaction amounts attached.
Procure-to-Pay triage
580 vendors analysed for risk patterns, outlier categories and flagged transactions.
ML model registry
19 trained model versions with training-sample counts up to 81k, feature counts and training durations — one active at a time. Real MLOps, not just LLM calls.
Ask AI for alerts
A chat assistant answers questions like 'show high-risk alerts and duplicates' with AI reasoning for policy violations.
By the numbers
/ PROOFQuestions, answered
FAQWhat is ICAT in Moderor?
ICAT (Smart Alert Triage) is the ML-driven triage product in the GRC Suite. It scores Travel & Expense and Procure-to-Pay activity for risk, flags outliers and duplicates, and explains policy violations with AI reasoning.
Does ICAT use real machine learning or just an LLM?
Real trained models. ICAT keeps a registry of 19 model versions with training-sample counts, feature counts and training durations, and reports false-positive reduction per model version. LLM-powered Ask AI sits on top for investigation.
How does ICAT reduce false positives?
Each retrained model version is benchmarked, and the dashboard reports an explicit 'FP Reduction By Model' metric, so triage precision is a tracked number — not a claim.
More from GRC Suite
/ RELATEDContinuous Compliance Monitoring
AI agents that evaluate every asset against every control, continuously.
View product →GRC SUITEAuditor Workbench
Plan, execute and track hundreds of audits with planned-vs-actual analytics.
View product →GRC SUITERisk Assessment
Vendor and third-party risk lifecycle with AI-maintained attention queues.
View product →See it on your data.
Connect a source over MCP, point an agent at a control set, and watch the first findings arrive — with you in command.