Software Engineer (Fraud Prevention)

Location: Hybrid in London or Belgrade

Company

Perplexity develops search products and infrastructure, including systems for fraud prevention and abuse detection.

What you will do

• Design, build, and operate high-throughput, low-latency backend services for fraud detection and prevention.
• Build data pipelines and real-time feature computation for detection systems and ML models.
• Collect and process trust signals, integrating fraud decisions into product, authentication, and payment flows.
• Own production reliability through observability, alerting, safe rollouts, and incident response.
• Build internal tooling for investigation, enforcement, and auditability.
• Partner with fraud analysts to convert emerging abuse patterns into production detections.

Requirements

• 5+ years of experience building and operating high-throughput, latency-sensitive backend systems in production.
• Strong systems programming skills; experience with Rust, Go, or C++ is preferred, with Rust used in the team.
• Hands-on experience with Kubernetes, Terraform, and cloud infrastructure.
• Solid data fundamentals, including stream and batch computations, caching, and storage trade-offs.
• A pragmatic, ownership-driven approach focused on incremental delivery, measurement, and safe failure degradation.

Nice to have

• Experience in fraud prevention, anti-abuse, or payments risk.
• Experience serving ML models in production.
• Client-side experience with web or mobile systems for signal collection.

Culture & Benefits

• Hybrid work in London or Belgrade.
• End-to-end ownership from system design through production operations.
• Close collaboration with fraud analysts, ML engineers, and product teams.