CodeMeet Madrid: Kubernetes Escape Room – Can You Debug Your Way Out with AI?
Detalles
# CodeMeet Madrid: Kubernetes Escape Room – Can You Debug Your Way Out with AI?
A service is down. Pods keep restarting. The logs aren't giving you the full picture. And you have AI as your copilot.
Can you find your way out?
On 29 October, we're heading to Nexthink's offices in Madrid for a hands-on CodeMeet focused on Kubernetes troubleshooting, incident response and AI-assisted debugging.
We'll start with an introductory talk from Manuel Fernández Arias, Consultant Software Engineer at Thoughtworks, exploring a question that is becoming increasingly relevant for engineers: Can we trust AI when production is on fire?
Then, we'll put those ideas into practice with Carlos Vicent and Unai Arrien, SREs at Nexthink, through an interactive Kubernetes Escape Room.
Rather than simply talking about how AI can help engineers, Carlos and Unai will put it into practice. You'll face realistic incident scenarios and use logs, metrics, events, recent changes and AI to investigate what is going wrong and figure out how to get the system back on track.
The goal isn't to let AI solve everything for us. It's about understanding where AI can make incident response faster, how to use it effectively and when human judgement is still essential.
[Register here]
Meet the speakers
Manuel Fernández Arias — Consultant Software Engineer at Thoughtworks
Manuel is a Software Engineer with more than six years of experience, currently working at Thoughtworks. He's interested in the intersection between software architecture and AI, particularly in how agents and LLMs can be integrated into real development workflows.
He'll kick off the evening with “Can We Trust AI When Production Is on Fire? From Stack Overflow to AI Agents”, exploring how AI is changing incident investigation and where human judgement and safety guardrails are still essential..
Carlos Vicent — SRE at Nexthink
Carlos leads reliability initiatives for more than 20 million endpoints across Nexthink's distributed platform, which includes 200+ microservices.
He specialises in incident response automation and observability for cloud-native environments, and is a big Kubernetes enthusiast. He also regularly attends DevOps, SRE and observability conferences around the world.
Together with Unai, he'll take us through the Kubernetes Escape Room and the different incidents we'll have to investigate.
Unai Arrien — SRE at Nexthink
Unai works on the operation and improvement of resilient distributed systems at Nexthink.
His experience covers automation, observability and incident response, and he particularly enjoys helping teams investigate, debug and solve complex problems from the real world.
Together with Carlos, he'll guide the hands-on part of the evening, where we'll investigate incidents, follow the clues and see what happens when AI becomes part of the debugging process.
Agenda
🕡 18:30 – 18:45 | Registration
👋 18:45 – 18:50 | Welcome
🎤 18:55 – 19:20 | Intro talk — Can We Trust AI When Production Is on Fire?
Manuel Fernández Arias will explore how AI is changing incident investigation, from searching Stack Overflow to working with AI agents directly from our terminals and clusters. We'll look at how to use AI as a copilot without blindly trusting its recommendations.
🧩 19:20 – 20:40 | Kubernetes Escape Room Workshop
Carlos Vicent and Unai Arrien, SREs at Nexthink, will lead us through an interactive Kubernetes Escape Room based on realistic incident scenarios.
Together, we'll investigate what’s going wrong, follow the clues and use logs, metrics, events and AI to diagnose each incident. Along the way, Carlos and Unai will share their approach to troubleshooting and show us where AI can speed things up — and where it can lead us in the wrong direction.
🍻 20:40 – 21:30 | Networking & closing
After surviving the Escape Room, it's time to relax, grab a drink and share the experience with the rest of the attendees.
What will we explore?
During the Kubernetes Escape Room, Carlos and Unai will take us through different incident scenarios inspired by real-world problems:
- How to investigate incidents in Kubernetes using logs, metrics and events.
- How AI can help us speed up troubleshooting and generate hypotheses.
- What information we need to give AI to get useful answers.
- How to validate AI recommendations before acting on them.
- What happens when AI gets it wrong or sends us down the wrong path.
- Where AI can help and where human judgement is still essential.
- How to move from a hypothesis to a safe mitigation and verify that the incident is actually resolved.
It will be less “let's talk about AI” and more “there's something broken — let's figure out what happened.”
Who is this for?
If you're interested in Kubernetes, SRE, DevOps, cloud, observability, incident response or applying AI to operations, this workshop is for you.
You don't need to be an AI expert. The workshop is about exploring how AI can become part of a real technical investigation and how engineers can work with it without blindly trusting what it says.
And yes, you'll actually have to escape the incident. 🔥
Before you join 👀
This will be a hands-on workshop, so you'll need to prepare a few tools and repositories before the event.
We'll send you everything you need in advance so you can get your environment ready and make the most of the workshop.
💻 Important: bring your laptop and make sure everything is set up before you arrive.
See you in Madrid
📅 29 October 2026
🕡 18:30 – 21:30
📍 Nexthink offices, Madrid
Free entry while places last
[Register here]