Reliability at Scale – With Bruno Paulino (N26)

🎧 New Señors @ Scale Episode

This week, I sat down with Bruno Paulino, Tech Lead at N26, to talk about what reliability at scale really means when every deploy touches millions of bank customers.

It’s a conversation about trust, resilience, and trade-offs — from building CI/CD pipelines that deploy in minutes, to server-driven UIs that let N26 roll out features instantly, to scaling customer support using LLMs and Retrieval-Augmented Generation.

⚙️ Main Takeaways

1. Reliability isn’t just uptime — it’s trust.
When your app handles real money, every minute of downtime has a financial cost. N26 designs reliability as a core product feature, not an afterthought.

2. CI/CD transformed how fast teams ship.
Pipeline time dropped from one hour to five minutes. With GitHub Actions, Argo CD, and Kubernetes, deployments are now predictable, testable, and safer.

3. Server-driven UI changes the game.
The backend defines what users see — from banners to buttons — across mobile, web, and desktop. Features can roll out in five minutes instead of waiting for app store approvals.

4. AI is reshaping customer support.
Bruno’s team built an AI Assistant using LLMs and RAG pipelines to make support faster, more contextual, and scalable — all while keeping reliability and compliance intact.

5. Safe experimentation with Statsig.
Every major feature at N26 goes through type-safe A/B testing using Statsig. Teams can ship to 5% of users, observe performance, and ramp up safely.

6. Reliability vs. Speed: the trade-off.
At N26, reliability always wins. “If your bank is down, someone might miss a mortgage payment.” The culture encourages testing, redundancy, and resilience at every level.

7. Security is everyone’s job.
Every team has a security champion responsible for audits and best practices. Vendor integrations go through a formal security assessment to protect customer data.

8. Lessons from outages.
A misconfigured CDN once broke CSS for every customer. That post-mortem led to stronger Terraform practices and better guardrails around infrastructure changes.

9. Accessibility and compliance are non-negotiable.
The team uses React Axe, Playwright, and Radix UI to ensure accessibility testing happens in CI — not as an afterthought.

10. Culture matters more than code.
From mentorship to lightweight RFCs, Bruno highlights how strong culture and communication make large systems — and teams — more resilient.

🧠 What I Learned

• Reliability isn’t about perfection — it’s about predictability under stress.
• Developer experience is a direct multiplier for reliability.
• Automation and experimentation make teams faster and safer.
• The best platform engineers think in trade-offs, not absolutes.

💬 Favorite Quotes

“If your bank is down, it’s not just a bug — someone might miss their mortgage payment.”
“We cut deploy time from one hour to five minutes — and people started shipping again.”
“Server-driven UI means we can change what users see in minutes, not days.”
“The goal isn’t to move fast — it’s to move safely, repeatedly, and confidently.”

🎯 Also in this Episode

• How N26 organizes its platform and assistance teams
• The hidden cost of slow pipelines and flaky tests
• Why CI/CD ownership belongs to everyone
• How accessibility became part of the release checklist
• The difference between reliability and availability

🎧 Listen Now

🎧 Spotify
📺 YouTube
🍏 Apple Podcasts

Episode Length: 1h 7m on reliability, platform culture, and CI/CD at scale.
If you care about reliability, developer velocity, or building systems people can trust — this one’s worth your time.

Happy scaling,
Dan