Systems | Development | Analytics | API | Testing

Recently someone on our team opened a traffic snapshot and found an HTTP request that was captured with our eBPF capture agent, nettap. Our protocol dissector parsed most of the response correctly, but that correctness ended once the response headers were processed. What they ended up with was a recording of an HTTP request/response where the response body was just an incorrect collection of garbage binary data when it should have been JSON text.

Every QA budget has a line item for building test coverage, but 30–50% of that automation budget ends up spent on maintenance instead of new tests. That disparity stays invisible until a release goes out, the application shifts underneath the tests, and the QA team spends the next three days rewriting broken scripts instead of finding new bugs.

I spent the day at NVIDIA thinking I’d be learning about all the newest models and RAM throughput achievements. Instead, a version of the future started to crystallize where AI really is good enough, pervasive and cheap(er). Let me share some of my observations and conclusions.

APIs are the backbone of modern software. They connect microservices, power mobile experiences, and make integrations possible across industries. For all their importance, API testing remains one of the most fragmented, manual, and maintenance-heavy parts of the software development lifecycle (SDLC). So as development accelerates in an AI-disrupted SDLC, application integrity – continuous, measurable assurance that your software just works as intended – becomes harder to maintain, not easier.

AI can now handle the slowest parts of test management, inside Jira. – The Zephyr Agent for Rovo creates test cases and links them to your work items, in the projects where your team already plans and builds. This guide covers how testing is changing in the AI age, what the agent is, where to find it, and how to run your first task.

Ask an AI for K-pop concert advice without saying the group, city, date, or budget. It may confidently send you to a BLACKPINK tribute night in Cleveland with a $400 resale ticket. The AI was plenty confident. It just had nothing real to go on. That is exactly what happens when developers test AI applications with invented traffic. The test may look reasonable. The result may even pass.

You've diagnosed the problem. Your agentic AI initiatives are stalling — not because the models are wrong, but because the integration layer underneath them wasn't built for this. Batch data, rigid schemas, fragmented governance, no real-time event delivery. Now the question is: what do you actually build, and how do you build it without tearing down the infrastructure you already have?