Trust & Safety

Designing and implementing machine learning-based detection systems for abuse, fraud, and policy violations

Analyzing attack patterns and emerging abuse trends to identify novel threat vectors and behavioral anomalies

Developing and maintaining enforceable policies, rules systems, and classification frameworks for platform safety

Conducting investigations into complex misuse cases involving suspicious user behavior and coordinated harm

Building automated response mechanisms and enforcement workflows that operate without manual intervention

Querying, transforming, and analyzing large datasets using SQL and data manipulation techniques

Writing and maintaining Python code for data analysis, system testing, and detection logic implementation

Translating ambiguous safety risks and complex threat landscapes into measurable, evidence-based problems

Building and operating systems to detect phishing, cryptomining, account takeovers, and financial fraud at scale

Scoping and implementing abuse monitoring systems for new product launches and existing platforms

Building dashboards, monitoring systems, and prevalence estimators for safety metrics and trends

Designing experiments and conducting causal inference analyses to understand safety intervention impacts

Building zero-to-one analytical systems and transforming prototypes into scalable, reusable tools

Designing threat taxonomies and harm classification frameworks for emerging and frontier risks

Performing high-volume content review and data labeling tasks with accuracy and attention to detail

Integrating and tuning security scanning tools in continuous integration pipelines

Designing large-language-model guardrails to detect abuse scenarios in AI-generated content and interactions

Detecting and analyzing agentic and autonomous behavior patterns in AI systems for safety risks

Using large language models as defensive tools to identify malicious patterns and automate threat classification

Conducting horizon scanning, competitive benchmarking, and external narrative analysis for risk sense-making

Applying prompt injection attack detection and mitigation techniques at production scale

Developing compliance programs aligned with global online safety and content moderation regulations

Conducting regulatory risk assessments and translating legal obligations into enforceable safeguards

Conducting behavioral and psychological analysis of user interactions with AI systems in high-risk contexts

Developing domain-specific expertise in chemical, biological, radiological, nuclear, and explosives threat detection

Communicating technical findings and risk assessments clearly to both technical and non-technical stakeholders

Coordinating cross-functional teams across Policy, Legal, Engineering, and Communications during high-stakes situations

Operating independently with high ownership in ambiguous, rapidly evolving problem domains

Identifying gaps in existing safety systems and proposing improvements based on investigation findings

Managing escalation procedures and on-call incident response operations for sensitive enforcement decisions