Building agentic quality into your delivery pipeline.

AI tools now write code faster than any team can test it, and that speed creates three compounding bottlenecks. More AI-generated code means more coverage gaps. More tests mean CI/CD pipelines that were never sized for the load. More failures mean triage that humans cannot keep up with. Each bottleneck feeds the next: build times balloon, time to resolution climbs, and trust in the green build erodes.

In this session, Ole Lensmar, CTO of Testkube, walks through a model where AI works across all three. Tests are generated and run inside the same platform that executes them. Agents prioritize and tune what runs in the pipeline. And agents take on triage, rerunning flakes, classifying failures, finding root cause, and opening issues automatically, so humans handle the exceptions rather than the queue.

You will leave knowing where AI genuinely accelerates quality across test creation, execution, and analysis, where it adds new risk, and how to put agents to work in your pipeline without sacrificing reliability or control.

What you will learn:

• The three compounding bottlenecks AI-velocity development creates in testing: creation, execution, and analysis


• Why those bottlenecks amplify each other, and the downstream cost in build times, MTTR, and release confidence


• How AI closes coverage gaps by generating and running tests in the same platform that executes them


• How agents orchestrate the pipeline with fail-first sequencing, sharding, and resource tuning to cut run time


• How agents triage at scale: auto-rerunning flakes, classifying failures, root-cause analysis, and opening issues so humans handle only the exceptions