Agent architecture
QAWave agents generate, heal, and triage test code through a reviewable CI/CD workflow. The goal is not a black box; the output is code your team can inspect.
Docs
Public technical notes for teams evaluating QAWave. Detailed implementation docs are shared during discovery because every stack is different.
CI/CD integration
Agents are designed to run inside existing pipelines such as GitHub Actions, GitLab CI, Jenkins, and CircleCI. No rip-and-replace is required for the first discovery scope.
Evals and governance
Every agent capability gets an eval plan: expected behavior, failure modes, approval gates, and audit trail. Thresholds are calibrated against your application before production use.
Security and data
Source code ownership stays with the customer. Data handling, DPA, subprocessors, and AI transparency notes are available in the legal section.