Partner with product, technical, and engineering stakeholders to align technical deliverables with roadmap milestones and support general-availability launches.
Own the end-to-end ML lifecycle for misdirected email, including data wrangling, feature engineering, model training, evaluation, deployment, and monitoring.
Run offline evaluations, online A/B tests, threshold-setting, targeted error analysis, and post-launch monitoring to improve model efficacy and prevent regressions.
Investigate production data and behavioral or trend shifts, and launch experiments that improve detection reliability and customer impact.
Participate in an on-call rotation for owned components, resolving efficacy-related alerts and investigating customer or internal reports of false positives and false negatives.
Maintain technical documentation and communicate effectively across distributed, cross-functional teams.
Requirements
Bachelor's degree in Computer Science, Machine Learning, Artificial Intelligence, Information Systems, or a related engineering or quantitative field.
At least 1 year of experience building and operating applied ML features in production systems.
Experience contributing to end-to-end ML systems, including data wrangling, feature engineering, model selection, training, evaluation, production deployment, and monitoring.
Ability to implement and reason about algorithms, develop features, combine signals, and apply numerical computing effectively.
Experience interrogating production data, identifying behavioral or trend shifts, and running targeted experiments to improve model efficacy.
Understanding of online and offline pipelines, data tables, and labeling workflows for safe and scalable model deployments.
Experience with offline metrics, online A/B tests, threshold-setting, drift and performance monitoring, guardrails, and rollback strategies.
Strong written and asynchronous communication skills and the ability to work independently across distributed, cross-functional teams.
Preferred experience with Python, Go, AWS, Spark, and Databricks.
Preferred experience in email security, data loss prevention, misdirected email prevention, customer-focused ML deployments, detector or rule development, and operationalizing research into reliable customer-facing systems.
Abnormal Security
Abnormal AI is the leading AI-native human behavior security platform, leveraging machine learning to stop sophisticated inbound attacks and detect compromised accounts across email and connected applications.