data analyst in antifraud

Higgsfield AI develops AI-powered video creation and next-generation creative tools.

• Run end-to-end ring investigations, from flagged accounts or cost anomalies to identifying clusters, members, signatures, lifetime economics, and recommended actions
• Build account linkages using payment instruments, IP and ASN concentration, email-stem families, registration bursts, behavioral signals, and automation fingerprints; distinguish evidence from coincidence
• Separate abuse from legitimate but unprofitable users, treating cost as a gate rather than a trigger
• Triage the detection system’s review queue and feed adjudication findings back into it
• Produce evidence packages that stand up to external auditors, payment networks, customers disputing bans, and Legal enforcing the Terms
• Write account-level findings when actions are account-level, including a reason for each ban
• Interpret the Terms of Use and flag when evidence supports a finding but not the proposed action
• Own and consistently define monthly abuse-loss reporting, net of reversed revenue and legitimate customers mistakenly caught
• Distinguish COGS burned, revenue at risk, revenue reversed, and cash refunded
• Build recurring reporting on abuse trends and provide same-day ad hoc analysis of compute-cost spikes
• Monitor detectors and identify rules that stop firing as detection events
• Verify data instruments before explaining changes, including upstream tables, joins, and vendor signals
• Identify emerging abuse patterns through cost anomalies, unusual cohorts, and support tickets
• Validate work built by the Antifraud data scientist and partner with Payments, Support, and Finance

• At least 2 years of investigative or risk analytics experience in fraud, abuse, trust and safety, AML, chargebacks, gaming, or marketplace integrity
• Strong SQL skills, including self-joins, window functions, cohort replay, and joining payment, usage, and identity data across messy keys
• Working knowledge of Python for analysis, including pandas, notebooks, and basic graph work
• Strong investigative instincts and the ability to distinguish meaningful patterns from coincidence
• Evidentiary discipline: write verifiable findings and distinguish suspected abuse from provable abuse
• Understand unit economics, including gross margin per account, compute costs, and why a negative-margin customer may be legitimate
• Know when to escalate rather than act on thin evidence, and document that judgment
• Clear written and spoken English at B2+
• Not a fit for candidates who want to build models full time, do content moderation or NSFW classification, ban accounts without clear reasons, treat every negative-margin account as abusive, require labeled datasets or clean tables to start, are uncomfortable with account restrictions that may sometimes be wrong, or want predictable 9–5 workdays
• Будет плюсом: experience as a fraud or risk analyst in fintech, payments, marketplaces, gaming, crypto, or betting; trust and safety investigation or platform integrity analysis; chargeback, dispute, or AML analysis with self-written queries; product or BI analysis experience involving abuse problems

• Competitive salary in USD
• Participation in the company's stock option program
• Relocation support, including a flight and temporary housing
• Paid on-site trial lasting 1 month
• Hiring process: first interview (30 min), business case (60 min), take-home with real-shaped data (7 days, ~10–12 hours of work), team interview (60 min), and paid on-site trial (1 month)
• Flat structure, high autonomy, and fast career growth