Data Scientist, Trust & Safety
Remote • Foster City, CA • FullTime
Posted 10d ago
About the job
Replit is seeking a Data Scientist to establish and grow its Trust & Safety and Anti-Abuse program. This role involves transforming complex data signals from user behavior, identity, payments, infrastructure, and content into effective measurement systems, detection mechanisms, and decision-making processes. The goal is to protect Replit's users, platform, and financial integrity by making abuse economically unviable while minimizing friction for legitimate users. The position is at the forefront of AI-native abuse, tackling threats like phishing, cryptomining, LLM token farming, and AI agent exploitation, requiring a proactive and adversarial approach to identify, measure, and respond to these evolving challenges.
Responsibilities
- Own the analytical foundation for Trust & Safety, including abuse prevalence, fraud loss, and detection/mitigation rates.
- Build reliable datasets and dbt models connecting various product and user signals.
- Develop and evaluate risk models, rules, and anomaly-detection systems for diverse threats.
- Design rigorous evaluations and tests to measure detection quality and policy impact.
- Define thresholds and decision frameworks balancing abuse reduction, economic loss, and user friction.
- Investigate emerging abuse patterns, quantify impact, and recommend solutions to product and engineering teams.
- Develop predictive models for account, device, and transaction risk.
- Partner with Support and Legal to enhance case review and feedback loops.
- Build monitoring systems to detect model drift, attacker adaptation, and data quality issues.
- Communicate findings, tradeoffs, and uncertainties to technical and non-technical partners.
Requirements
- 5+ years of experience in data science, product analytics, fraud, risk, trust and safety, or a related field.
- Strong SQL and Python skills.
- Experience working with large behavioral datasets.
- Experience building reliable data models.