Inside Revolut Engineering: Banking at Hyperscale

  ____  ____  ____  ____ 
 / ___|/ ___||  _ \| __ )
 \___ \\___ \| |_) |  _ \
  ___) |___) |  _ <| |_) |
 |____/|____/|_| \_\____/
Revolut - Tech Stack 2025

Revolut processes $2T+ payments yearly with <50ms latency. How?

Tech Stack Highlights

Infrastructure: Kubernetes + Finagle

// Finagle service (Scala)
class PaymentService extends Service[Req, Resp] {
  def apply(req: Req): Future[Resp] = {
    for {
      validated <- validator(req)
      charged <- stripe.charge(validated)
      notified <- notification.send(charged)
    } yield SuccessResp
  }
}

Data Platform

• Kafka Streams for real-time fraud
• ClickHouse for analytics (1PB+)
• Vitess for MySQL sharding

Engineering Culture

Weekly Tech Talks:

• Rust for payments infra
• WASM for micro-frontends

Open Source:

• REngine – Kotlin coroutines framework

Challenges Solved

1. 99.999% Uptime: Chaos engineering with Litmus
2. Regulatory Compliance: GDPR + PCI-DSS automation
3. Multi-Cloud: GCP + AWS hybrid

Key Lesson: Monolith → Microservices → Platform Engineering.

Careers @ Revolut Tech

• 400+ engineers, 50 nationalities
• Remote-first (post-COVID policy)
• Stack: Scala, Kotlin, Go, Rust

Follow @RevolutEngineering for deep dives.

[Data from Revolut Tech Blog 2024]