Architecture & CloudConference50min
How We Failed Before We Succeeded in Building a High-Value Payment System
A practical story of building an RTGS payment system, covering failed early designs and the architecture that worked in production. It highlights challenges in real-time settlement, consistency, regulation, liquidity management, netting, and gridlock resolution, and explains why standard distributed systems patterns often needed rethinking.
talk.summaryAiDisclaimer
Badr BellajMchain
Building a high-value payment system turned out to be much harder than we expected. We quickly learned that processing payments isn't just about moving money from one account to another. Every architectural decision has consequences when transactions must settle in real time, remain consistent under failure, and meet strict regulatory requirements.
In this session, I'll share our journey of building an RTGS (Real-Time Gross Settlement) system—from the early designs that didn't work to the architecture we finally trusted in production. We'll explore the engineering challenges behind payment queues, liquidity management, netting, and gridlock resolution, and discuss why many common distributed systems patterns had to be reconsidered in this context.
This isn't a talk about the "perfect" architecture. It's a story of wrong assumptions, hard lessons, and the practical decisions that helped us build a resilient payment system. If you've ever wondered what it really takes to engineer critical financial infrastructure where every transaction counts, this talk is for you.
In this session, I'll share our journey of building an RTGS (Real-Time Gross Settlement) system—from the early designs that didn't work to the architecture we finally trusted in production. We'll explore the engineering challenges behind payment queues, liquidity management, netting, and gridlock resolution, and discuss why many common distributed systems patterns had to be reconsidered in this context.
This isn't a talk about the "perfect" architecture. It's a story of wrong assumptions, hard lessons, and the practical decisions that helped us build a resilient payment system. If you've ever wondered what it really takes to engineer critical financial infrastructure where every transaction counts, this talk is for you.
Badr Bellaj
Badr is CTO at Mchain and has over 18 years of experience designing and building distributed systems. He holds a PhD in Distributed Systems Security from Institut Polytechnique de Paris and is the author of the book Blockchain by Example. His work focuses on blockchain, cybersecurity, digital trust, and emerging technologies, helping organizations design secure and scalable distributed platforms.