Contract Risk Evaluation Specialist

The Risk Assessor Agent employs GPT-4 Turbo with low temperature (0.2) for analytical precision. It evaluates contract risks using logistic regression scoring models, calculates financial exposure using Monte Carlo simulation (1000 iterations) with 95% confidence intervals, performs trend analysis to identify risk patterns across document portfolios, detects anomalies that may indicate hidden risks, and categorizes risks by type (Liability, Financial, Compliance, Operational, Reputational, Legal, Data Privacy) and severity (Critical, High, Medium, Low, Info). The agent generates detailed findings with evidence, impact assessment (financial min/max estimates, legal implications, operational considerations, timeline urgency), and actionable recommendations. It triggers HITL approval requests for critical findings exceeding defined thresholds, ensuring human oversight for high-stakes decisions. Tools include risk_scorer (800ms avg, 97% success), financial_modeler, monte_carlo_simulator (2.5s avg, 98% success), and anomaly_detector.