How we've evolved our CI/CD in the agentic era

SageOx has repeatedly redesigned its CI/CD, pull request, and deployment workflows to accommodate agent-assisted development, where engineers and agents generate high volumes of often very large changes. The company moved away from traditional human-driven practices such as manual git bisecting toward agentic debugging, extensive automated testing, and a system in which agents continuously identify customer issues and bugs, file tasks, and sometimes implement low-risk fixes. To reduce lengthy CI/CD delays and infrastructure constraints, it parallelized and sharded tests, optimized GitHub Actions runners, introduced local automated code-review checks, and adopted merge queues to let developers finish sessions without waiting for merges. Its workflow now separates faster PR validation from merge-queue checks and runs the most expensive end-to-end and behavior-driven tests periodically after changes reach main rather than on every merge, accepting limited risk in exchange for faster delivery to test environments. The approach reflects a shift from optimizing primarily for strict per-merge validation to optimizing for rapid feedback, continuous testing, and the ability for teams to experience new features quickly as agentic development accelerates software output.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.