Operations | Monitoring | ITSM | DevOps | Cloud

AI didn’t just change how we write code. It changed everything that comes after. Application teams are shipping more code than ever with AI — but 70% of the work still happens after coding: testing, security, deployment, optimization, and keeping everything moving. As coding gets faster, delivery becomes the bottleneck. That’s where Harness comes in.

In this special ShipTalk episode, host Dewan Ahmed (Principal Developer Advocate, Harness) sits down with @Harnessio Field CTO Nick Durkin for spicy—but practical—2026 predictions across AI, software delivery, DevSecOps, MLOps, and developer experience. Will we see the first “AI-caused meltdown”? Are AI “confidence scores” even trustworthy? Is 2026 the year of AI cleanup crews and recovery engineering? Nick’s take: the answer isn’t more gates—it’s guardrails, policy in the pipeline, and teams operating with the same “rulebook.”

Message brokers are a critical component of modern distributed systems, facilitating asynchronous communication between services. Load testing message broker integrations requires special considerations since the interaction patterns differ from traditional HTTP-based APIs. Speedscale provides specialized tooling to help you load test applications that integrate with message brokers by.

OpenTelemetry aims to link metrics to traces and logs, offering OpenCensus users a seamless migration path. Work with existing protocols like Prometheus. Leverage existing tooling without learning something completely new.

OpenTelemetry explained: standards, SDKs for various languages (Ruby, Python, Go), and middleware tools. Deploy these to pre-process data and send it to your destination.

Platform engineering provides powerful tools that handle a lot under the hood. Learn how to calculate your remaining error budget with a simple formula using real numbers and objective statements.