Harness

Harness is a CI/CD and software delivery platform with modules for builds, deploys, feature flags, cloud cost and security testing; paid plans are quoted.

Harness takes a code change from the pull request through build, test, security checks and deployment, then follows what the running service costs in the cloud. Instead of a separate CI server, deploy tool and flag service, a team runs all of them from one account and one pipeline model. It comes from Harness Inc., an independent company started in 2017, with its headquarters in San Francisco. Continuous Integration runs builds on Harness Cloud machines with Linux, macOS and Windows images, or on a team's own VMs and Kubernetes clusters. Cache Intelligence reuses dependencies between runs, and the test reports flag tests that pass and fail inconsistently. Continuous Delivery and GitOps deploy through Delegates on the team's own infrastructure, with Harness approval stages, permission-guarded production environments and deployment freeze windows. Feature Management and Experimentation controls rollouts, Cloud Cost tracks spend, and Security Testing and Chaos Engineering, built on Litmus, probe the result. Infrastructure as Code Management, Artifact Registry and Code Repository with its Merge Queue round out the modules, while Git Experience keeps pipelines in Git and the Template Library shares steps and stages across projects. Harness is built for platform engineering and DevOps teams that run delivery for many developers, and for enterprises that need approvals and Policy as Code around production. Individual developers and small teams can start on Harness Open Source with community support. Growing organizations with small to mid-size teams take the Essentials bundle of CI, CD, IaC and security. Larger companies choose modules one by one on Enterprise, with premier support and an account team. A typical pipeline starts from a push or pull request on GitHub, GitLab, Bitbucket, Azure Repos or Harness Code. It builds and tests the change, then deploys it to Kubernetes, AWS, Google Cloud or Azure. Jobs reach those clouds with short-lived OIDC credentials instead of stored keys, and secret values are masked in the logs. Cron triggers start pipelines on a schedule as well, and teams that track releases in Jira and ServiceNow connect both to the same platform. Pipeline Insights dashboards show how long each pipeline, stage and step takes and how often it succeeds, which is how platform teams find the slow and failing parts of their delivery.