Forward Deployed Engineer, Trust and Safety

About the Team

We’re people that are passionate about making the internet a safer and more trusted place for all. We love the fraud and trust & safety space and want to teach companies how they can protect themselves, their users, and create frictionless experiences for legitimate consumers.

Responsibilities

• Work with our Trust and Safety Architect and Data Science teams to surface emerging fraud patterns across the network, escalate, and proactively take them down.
• Detect patterns and turn those findings into sharper signals, tighter configurations, and smarter decisioning logic.
• Work across different verticals and closely with customers, partners, and prospects with different risk appetites—some optimizing for approval rates, some minimizing chargebacks, some fighting account takeover and other types of abuse.
• Help build dashboards, tune and build models, decision logic, and custom signals to help customers achieve their desired business outcomes.
• Identify sources of false positives, possible coverage gaps, and other vulnerabilities by digging into raw event streams; form a hypothesis, design a test, and implement the fix.
• Lead forensic investigations during fraud spikes: trace attack patterns to their source, identify the technique being used, deliver a clear writeup with remediation steps.
• Distinguish between one-off anomalies and systemic gaps that indicate a product opportunity—and advocate for the latter with rigor.
• Contribute to detection frameworks, investigative tooling, and internal playbooks that make every engineer and analyst at Sift more effective.
• Be the conduit between customer reality and internal roadmap; your field observations should directly accelerate what Sift ships next.
• Some travel may be required.

Requirements

• 5–8 years in fraud, trust & safety, risk, or a closely related data science domain—you've spent meaningful time working with fraud data, not just adjacent to it.
• Strong SQL and Python skills; you reach for code to answer a question, not to build a pipeline.
• Strong understanding of ML concepts applied to fraud: classification models, feature engineering, precision/recall tradeoffs, threshold calibration, score drift.
• Experience analyzing large-scale behavioral or transactional datasets to find patterns and anomalies—you know what a fraud ring looks like in the data, not just in a textbook.
• Ability to communicate technical findings to both technical and non-technical stakeholders; you can write a forensic investigation report and present it to a VP of Risk in the same week.
• Customer-facing experience; you understand that different businesses have different priorities, and that listening before optimizing is part of the job.

Nice to Have

• Hands-on experience with fraud detection platforms (in-house or 3rd party).
• Hands-on experience building with AI: LLM APIs, prompt engineering, or agentic workflows—whether that's automating an investigation step, building a tool that surfaces patterns from raw data, or wiring together a multi-step agent to accelerate fraud analysis.
• Familiarity with real-time event processing systems.
• Experience with rules-based decisioning systems alongside ML—knowing when a hard rule beats a model score.
• Background in payments, e-commerce, fintech, marketplace, or account security fraud.
• Prior forward deployed, staff engineering, or embedded consulting experience at a technical product company.
• Computer Science, Data Science, Mathematics, Statistics, Information Systems, Economics degree or equivalent.