What tech stack does Stripe use?

Stripe's backend is famously built on Ruby — reportedly one of the largest Ruby codebases in existence, on the order of tens of millions of lines. To manage that scale, Stripe built and open-sourced Sorbet, a static type checker that runs across the entire Ruby codebase. Newer and performance-sensitive services increasingly use Java and Go.

Infrastructure runs on AWS, which gives Stripe the elastic capacity to absorb spikes in payment volume across global regions. The platform leans on event-streaming (Kafka) and container orchestration, with a continuous-integration system that runs tens of thousands of test suites to keep a payments-critical codebase safe to ship.

On the front end, Stripe engineers work primarily in TypeScript and React — the same stack behind its Dashboard, Checkout, and documentation experiences, which are themselves a competitive advantage. Mobile SDKs are built natively in Swift (iOS) and Kotlin (Android).

On the data side, teams use Scala and Apache Spark for large-scale processing, and Stripe runs substantial machine-learning infrastructure — most visibly in Radar, its fraud product, and increasingly across AI-driven and agentic-commerce features. This ML investment is a long-standing part of Stripe's engineering identity, not a recent bolt-on.

Stripe is a sophisticated, build-heavy engineering organization that has authored its own core tooling (Sorbet) and runs a massive AWS footprint — so the displacement/integration pitch matters. Vendors that complement (rather than try to replace) an AWS-and-open-source core — observability, data infrastructure, security, ML/AI tooling, developer productivity — map best; pitching to rip out core systems will not land.

Because Stripe has a strong build-vs-buy bias and deep internal platform teams, the realistic wedge is tooling that saves engineering time at scale or covers domains outside its core (e.g., specialized security, compliance, or analytics). Note that this stack is inferred from public signals — confirm specifics with the relevant team before anchoring a pitch on any single technology.