Overview In this role, you apply statistical modeling and data science techniques to transaction and telemetry data to identify anomalous traffic and behaviors. You'll build baselines and anomaly detection models using time-series analysis, clustering, correlation, outlier detection, and thresholds. You will engineer features from multiple data sources and develop scoring and classification models to differentiate legitimate, duplicate, retry, anomalous, and potential DDoS activity. You'll validate, monitor, and tune models, and work with engineering to productionize solutions while providing explainable, evidence-based recommendations to investigators.
Responsibilities- Develop baselines and anomaly detection models using time-series, clustering, correlation, and outlier detection
- Engineer features from transaction, connection, velocity, retry, batch and historical alert data
- Create statistical scoring and classification models for multiple traffic categories
- Validate models with historical and simulated traffic, evaluate precision, recall, FPR/FNR, and latency
- Monitor model performance and drift, fine-tuning detection algorithms
- Collaborate with engineering/operations to productionize models and provide explainable recommendations
- Integrate detection signals into agentic AI investigation workflows for automated correlation and evidence generation
- Analyze historical incidents to optimize thresholds and reduce false positives
Key requirements- data-driven decision making
- collaboration with cross-functional teams
- clear communication of statistical concepts to investigators
- time-series analysis
- anomaly detection
- feature engineering
