From 0 to €500 Billion and Beyond, How Adyen Builds for Growth
summary
Adyen describes engineering that scales from startup to handling half a trillion processed last year, over 80 million transactions per day across 12 data centers and 100+ TB Postgres clusters. The platform uses a monorepo, weekly release trains, and in-house pieces — a custom RPC/web framework, a Postgres-based queue and streaming system built for ordering and exactly-once semantics since 2016. A 32 TB table limit is encountered and patched without interrupting synchronous payment authorization.
AI-written summary · about
chapters (23)
- 0:00Intro
- 0:44Introduction
- 1:20Agenda
- 2:06Who is Adyen
- 2:30How does Adyen work
- 2:50Payments 101
- 6:08What makes this landscape so complex
- 7:53The Adyen Platform
- 9:11Our Own Data Centers
- 10:32Payment Flow Architecture
- 12:48Synchronous Payments
- 15:01Asynchronous Payments
- 16:40Q Implementation
- 18:53Payment Journey
- 21:39Postgres Limits
- 23:32Approach to Technology
- 25:01Technology Curve
- 26:37Technology Radar
- 27:49Scaling
- 28:34Monorepo
- 29:43Interview Questions
- 31:30Technical Organization
- 32:39Outro
related talks
10mAdyen’s journey from start-up to IPO
23mAdyen on the rise
20mCollision 2015 - Roelant Prins of Adyen with Dominic Rushe of The Guardian
50mPostgreSQL Scale-Out at inMobi
19mDemystifying the Data Stack of the Largest and Fastest Growing Payment Gateway in India
28mOrchestrating payments for the millions
25mScaling AI for high-stakes, real-time payments
54mBreakout Session