CodiLime at AutoCon 4 – scaling with eBay and AI troubleshooting

Date Published

CodiLime at AutoCon 4 – scaling with eBay and AI troubleshooting

AutoCon 4 in Austin, Texas, brought together the industry’s best for almost a week of deep-dive technical sessions, a sold-out workshop, and countless conversations with industry leaders. Last week, the CodiLime team was at the heart of this shift towards intelligent automation.

A highlight of our presence was the presentation by CodiLime’s Senior Software Engineer, Katarzyna Kurowska, and eBay’s Director of Site Network Engineering & Security, Aldrin Isaac.

Their talk, “Spectron: network orchestration in practice at scale – what works, what hurts”, offered a candid look behind the scenes of building an intent-based network automation platform. Instead of theoretical models, the session focused on hard-earned lessons: the importance of smart multi-layer abstractions, and the necessity of continuous network lifecycle support with digital twin capability. This approach is crucial for simplifying management in a multi-vendor heterogeneous environment, helping organizations optimize TCO (total cost of ownership) and ensure product reliability.

Moving from theory to practice, our experts, Adam Kułagowski and Tomasz Janaszka, led a sold-out workshop titled “From Chaos to Clarity: AI-Driven Network Troubleshooting.” Participants were challenged to diagnose failures in a live 200+ node network environment. The goal? To address one of the major NetOps challenges – reducing MTTR (mean time to repair). Instead of treating AI as a guaranteed fix, the session was a reality check allowing engineers to test the technology firsthand and check if it truly accelerates RCA (root cause analysis) in complex topologies.

Read the full deep-dive analysis

The insights gathered in Austin go far beyond what we can cover in a single news update. Our expert on the ground, Tomasz Janaszka, has published a detailed article where he described the major trends from the conference, including the evolution of intent-based networking, and analyzed them through the lens of our project experience.

You can read the full article here.