Data Scientist, Trust & Safety

Seeking a data scientist to design and scale a Trust & Safety and anti-abuse program from the ground up, onsite in Foster City, CA.

Responsibilities

• Own the analytical foundation for Trust & Safety, including abuse prevalence, fraud loss, false-positive/false-negative rates, time to detect, time to mitigate, appeal and reversal rates, and verification step-up conversion.
• Build reliable datasets and dbt models linking product events, identity signals, payment activity, infrastructure usage, content classifications, enforcement actions, appeals, and support outcomes.
• Develop and evaluate risk models, rules, and anomaly-detection systems for threats such as phishing, scam hosting, cryptomining, token farming, payment fraud, promotional abuse, and AI-agent exploitation.
• Design rigorous offline evaluations, shadow-mode tests, holdouts, and controlled experiments to measure detection quality and user impact of new policies and enforcement actions.
• Define thresholds and decision frameworks balancing abuse reduction, economic loss, customer friction, and false positives across free, paid, and enterprise users.
• Investigate emerging abuse patterns, quantify impact, identify coordinated behavior, and translate signals into clear recommendations for product and engineering teams.
• Develop predictive risk signals for accounts, devices, transactions, workspaces, or deployments and embed them into detection, review, and escalation workflows.
• Partner with Support and Legal to improve case review, appeals, reason-code quality, and feedback loops so human decisions inform model and policy signals.
• Build monitoring for drift and abuse adaptation, data-quality issues, and unintended harm to legitimate users.
• Communicate findings clearly to technical and non-technical partners, outlining tradeoffs, uncertainty, and evidence behind high-impact decisions.

Requirements

• 5+ years of experience in data science, product analytics, fraud, risk, trust and safety, or related fields.
• Strong SQL and Python skills with experience handling large behavioral datasets and building reliable data models or pipelines.
• Experience developing and evaluating predictive models, experiments, or decision systems, with sound judgment around uncertainty and tradeoffs.
• Ability to turn ambiguous data into clear recommendations and communicate effectively across technical and non-technical teams.
• Comfort with imperfect labels, biased samples, and high-impact decisions where false positives matter.
• Proficiency using AI tools to increase effectiveness while upholding analytical quality.

Technologies

• dbt, BigQuery, Snowflake, Fivetran, Amplitude, Mixpanel, Segment, Python, SQL

Benefits

• Competitive Salary & Equity
• 401(k) Program with a 4% match (US Only)
• Health, Dental, Vision and Life Insurance
• Short Term and Long Term Disability
• Paid Parental, Medical, Caregiver Leave
• Flexible Time Off (FTO) + Holidays
• Commuter Benefits (In-Office Only)
• Monthly Wellness Stipend
• Autonomous Work Environment
• In Office Set-Up Reimbursement (In-Office Only)
• Quarterly Team Gatherings
• In Office Amenities (In-Office Only)

Who You Are

• You're a data scientist who moves fast, goes deep, and thinks adversarially, able to spin up analyses in hours rather than days without compromising rigor.
• You investigate beyond top-line abuse metrics to understand selection effects, missing labels, policy changes, attacker adaptation, and hidden false positives.
• You recognize that Trust & Safety data is imperfect and outcomes are high-stakes, with delayed or biased ground truth and evolving attacker behaviors; you stress-test work and distinguish correlation from the evidence needed for enforcement.
• You leverage AI tools to amplify output—writing code, exploring data, generating hypotheses, and prototyping investigations—yet treat AI-assisted results as drafts and uphold a high standard for deliverables.

Examples of What You Could Do

• Create a measurement framework for abuse exposure, reconcile incomplete labels across automated detections, human reviews, appeals, chargebacks, and support cases, and establish a trustworthy baseline.
• Design a risk-scoring model for suspicious account clusters using identity, device, payment, graph, and product-behavior signals; set thresholds to reduce fraud while protecting legitimate users.
• Review a phishing rule that appears precise but disproportionately flags paying users with legitimate brand references; redesign its evaluation and review path to lower false positives.
• Quantify a progressive verification ladder, deciding when to escalate verification and weighing abuse prevention against legitimate-user conversion.
• Detect coordinated token-farming or promotional abuse networks by merging account-linkage graphs, referral data, payment patterns, and infrastructure usage; collaborate with Engineering to operationalize findings.
• Evaluate a new enforcement policy in shadow mode, estimate counterfactual impact, and advise launch, revision, or rejection prior to user exposure.

Preferred Qualifications

• Experience building or evaluating anti-abuse, fraud, identity, security, spam, integrity, or content-safety systems at scale.
• Developed, shipped, and maintained ML models in production (classification, anomaly detection, or risk scoring), including feature engineering on behavioral/transaction data, thresholding, and post-launch monitoring.
• Experience with graph analysis, entity resolution, coordinated-behavior detection, reputation systems, anomaly detection, or risk scoring.
• Experience measuring false positives and enforcement harm, designing human-review workflows, or using appeals and case outcomes as model feedback.
• Familiarity with progressive verification, KYC, account trust, or identity providers such as Prove, Persona, Socure, or Stripe Identity.
• Experience with causal inference methods such as difference-in-differences, propensity score methods, synthetic control, or uplift modeling.
• Experience with a modern data stack such as dbt, BigQuery, Snowflake, Fivetran, Amplitude, Mixpanel, or Segment.
• Experience at a consumer platform, developer tool, cloud provider, marketplace, fintech, or other product with meaningful adversarial exposure.

Bonus Points

• Built AI-powered analytical tools, investigation systems, automated detections, or novel measurement approaches.
• Experience with AI-native abuse such as prompt injection, LLM token farming, model extraction, or agent-driven abuse.
• Understanding of freemium, usage-based, or promotional pricing and the abuse incentives they create.
• Worked directly with operational review teams and can translate analytical signals into practical playbooks, queues, and escalation paths.

Want to Learn More About What We Are Up To

• Meet the Replit Agent
• Replit: Make an app for that
• Replit Blog
• Amjad TED Talk

Interviewing + Culture at Replit

• Operating Principles
• Reasons not to work at Replit