Backend Engineer; Node.js Risk

• Remote
• Full Time
• Node.js , Redis , SQL , Type Script

Our client provides high-limit credit cards to Private Banking clients who treat $100k transactions as routine. Our client needs a Senior Engineer to build the Risk & Dispute Defense Layer of their platform.

You will build the Node.js microservices that sit between our clients and the Visa network. Your code must make split-second decisions to approve a luxury purchase or block a BIN attack, and automate the heavy lifting of Visa dispute resolution.

• Event-Driven Architecture: Architect an event-driven system (using Kafka or Rabbit
  
  MQ) to ingest transaction webhooks.
• Velocity Logic: Write efficient, non-blocking Node.js logic to detect complex patterns (e.g., "72 charges in 48 hours" or "card testing attacks") using Redis for atomic counting.
• Latency Management: Ensure all fraud checks complete in under 200ms to avoid timing out the Visa ISO 8583 message flow.
• Visa API Integration: Build robust integrations with Visa/Processor APIs (VROL) to automate the submission of chargebacks and retrieval requests.
• Evidence Gathering: Create a service that automatically pulls transaction logs, IP history, and user activity to generate PDF evidence packages for disputes, reducing manual operations.

• Idempotency: Implement strict idempotency keys to ensure they never double-charge a client or double-submit a dispute, even if the network fails.
• Database Design: Optimize SQL schemas (Postgres) to store millions of transaction logs for instant querying during fraud investigations.

• Tech Stack: 5+ years in Backend Engineering, with expert proficiency in Node.js & Type Script.
• Industry
  
  Experience:
  
  3+ years in Fin Tech, Card Issuing, or Acquiring. You must know what a “reversal” is and how it differs from a “refund.”
• Performance: Experience tuning Node.js for high-throughput/low-latency (e.g., event loops, garbage collection tuning).
• Risk IQ: You can distinguish between a “Friendly Fraud” chargeback and a “Criminal Fraud” attack and know how to code rules for both.