Description
Own fraud forensics for the identity + verification platform (phone-number and OTP verification). Investigate attacks,
triage patterns, turn findings into rules and monitoring. Investigate SMS pumping campaigns + sophisticated fraud
attacks. Tune anti-fraud thresholds (every rule change trades blocked attackers against legitimate users lost at signup).
Own this space end to end while a permanent team member is hired. Leave behind detection the team can run without
you.
Must have
• 5+ yrs analytics engineering, DS or fraud analytics; 2+ yrs hands-on fraud or abuse detection
• Expert SQL on BigQuery - large event tables, window functions, partition-aware, cost-efficient queries
• Strong Python (pandas, notebooks) for ad-hoc forensics + pattern mining
• Scheduled pipelines in Airflow
• Dashboards in Looker (or equivalent BI)
• Tune rule- or threshold-based fraud controls incl. backtesting + false-positive analysis
• Works independently from day one; crisp communication to engineers + leadership
Nice to have
• Telecom / SMS fraud at scale (SMS pumping, IRSF)
• CPaaS provider logs + billing (Vonage, Twilio, Sinch, Infobip)
• Identity/verification: OTP funnels, rate limiting, velocity rules, device attestation (Play Integrity, App Attest)
• Fraud in emerging markets (Africa, South Asia) - different carrier/pricing patterns
• Event streams from Kafka; streaming anomaly detection (Dataflow/Flink)
• Statistical anomaly detection or lightweight ML (BigQuery ML, isolation forests)
Skills
Fraud Analytics
Fraud Forensics
SQL
BigQuery
Python
pandas
Airflow
Looker
Pattern Mining
Threshold Tuning
Backtesting
OTP
SMS Fraud
CPaaS
Kafka
Dataflow
Flink
Anomaly Detection
BigQuery ML
