Agentic development and deployment

  • Release version: Zurich
  • Updated June 5, 2026
  • 3 minutes to read
    • Summarize
    Summarized using AI
    This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.

    Summary of Agentic development and deployment

    The ServiceNow AI Platform streamlines app development and deployment through an integrated agentic workflow. After generating and expanding an app, the deployment process involves collaborative design, AI-driven development, thorough review and testing, developer validation, deployment approvals, and autonomous AI scans for issues before final deployment. This ensures secure, efficient, and compliant app delivery.

    Show full answer Show less

    Deployment Workflow

    • Collaborative design: Business and IT teams collaborate on requirements using preferred tools.
    • AI-driven app development: Build Agent and Now Assist generate and implement app updates based on process files, chat history, and diagrams.
    • Review and testing: Teams preview and revise updates, performing performance and readiness tests.
    • Developer review: Developers compare versions and approve AI-generated changes for deployment readiness.
    • Deployment approval: Deployment managers initiate approval workflows.
    • Autonomous checks: AI agents detect and remediate issues like sensitive data exposure before deployment.
    • Final deployment: Approved apps are deployed securely and efficiently.

    Deployment Options

    ServiceNow supports multiple deployment methods for agentically developed apps:

    • Git-based source control integration: Push scoped apps to Git repositories (e.g., GitHub, Bitbucket) via ServiceNow IDE for version control, branching, merging, and CI/CD automation.
    • Update sets and application packaging: Use System Update Sets to track changes and package them into scoped applications for easier transport across instances, including publishing to Application Repository (AppRepo).

    Moving Apps Through Environments

    • Wrap scoped applications in update sets and convert to AppRepo for deployment requests or standard update set processes.
    • Publish apps to AppRepo using Git or update sets to distribute across environments and integrate with ReleaseOps pipelines, which can automatically run Automated Test Framework (ATF) tests.
    • Register and entitle apps before publishing.

    Additional Deployment Tools and Capabilities

    • App Engine Management Center (AEMC): Conducts governance checks (ACLs, roles, compliance) before deployment, provides dashboards for monitoring, and manages app versions throughout the lifecycle.
    • ReleaseOps: Facilitates moving changes through multiple instances using customizable playbooks, automates update set validation and ATF testing, and manages approvals and scheduled releases.
    • ServiceNow SDK: Enables moving applications between local machines and instances via CLI, supporting integration with external CI/CD pipelines.
    • Automated Test Framework (ATF): Allows generation and execution of tests to verify app functionality after changes, integrated with ServiceNow Studio or IDE.

    This comprehensive deployment integration empowers ServiceNow customers to efficiently develop, test, validate, and deploy AI-generated applications with governance, automation, and quality assurance built in.

    After you generate and expand an app on the ServiceNow AI Platform, the deployment process is tightly integrated into the workflow.

    Workflow for deploying an agentically developed app

    After development, review, and testing are complete, a typical deployment workflow includes the following steps:
    1. Collaborative design: Business owners and IT collaborate on requirements and ideas using their preferred tools.
    2. AI-driven app development: Build Agent and Now Assist process files, chat history, and diagrams to generate and implement app updates.
    3. Review and testing: Teams preview updates, make revisions, and run rounds of performance and readiness testing.
    4. Developer review: A developer reviews the AI-generated changes, compares versions, and confirms the changes are ready for deployment.
    5. Deployment approval: The project is handed off to a deployment manager, who initiates the deployment approval process.
    6. Autonomous checks: AI agents automatically scan for issues (such as sensitive data exposure or model integrity problems) and remediate them before deployment.
    7. Final deployment: After all readiness scans and approvals, the new app is deployed securely and efficiently.

    Deployment options

    Build Agent supports the following deployment methods for apps created and edited with agentic development:
    • Git-based source control integration with ServiceNow IDE: ServiceNow supports Git-based workflows for version control and CI/CD.
      • You can push scoped apps to Git repositories, enabling branching, merging, and automated deployments. ServiceNow IDE supports bring-your-own Git integration, such as GitHub or Bitbucket.
      • For more information, see Integrating source control with the ServiceNow IDE.
    • Update sets and application packaging: Standard ServiceNow deployment uses System Update Sets to track changes.
      • Advanced guidance includes packing update sets into scoped applications for easier transport and installation across instances, for example using Application Repository (AppRepo).
      • For more information on System Update Sets, see System update sets.

    Workflows for moving apps through higher instances

    After you create an app using Build Agent, you have several options to move the app to the test instance.
    1. Wrap the entire scoped application in an update set. The workflow is as follows:
      1. Go to the Custom Applications list, select an app and swap to its scope.
      2. Convert the app to AppRepo.
      3. Publish the update set with demo data.
      4. Put the update set in a deployment request for ReleaseOps, or follow your standard update set process for deployment.
    2. Publish the app to AppRepo:
      • You can use a Git-based process or update sets to publish to AppRepo.
      • Scoped apps, as well as apps that are ready for testing, can be published to the AppRepo for distribution across environments.
      • After an app is in AppRepo, you can move it through a ReleaseOps pipeline. If ATF tests are included in the pipeline, they automatically run.
      • Register and entitle apps before publishing.
      • For more information on Application Repository, see ServiceNow application repository.

    Additional deployment tools and capabilities

    The ServiceNow AI Platform has additional deployment tools that include the following tools:
    • App Engine Management Center (AEMC):
      • After developing an app, submit it to AEMC for governance checks.
      • AEMC validates ACLs, roles, and compliance settings before deployment.
      • Use ReleaseOps pipelines to move apps through environments with ATF tests and approval gates.
      • AEMC provides dashboards for monitoring deployments and managing app versions throughout the lifecycle.
      • For more information on AEMC, see Using the App Engine Management Center.
    • ReleaseOps:
      • Move changes from development to production through multiple instances using customizable playbooks.
      • Automate preview, commit, and validation of update sets before deployment.
      • Run Automated Test Framework (ATF) tests as part of the pipeline to validate quality.
      • Deploy changes immediately or schedule releases for controlled rollouts.
      • Enforce checks, scans, and approvals before production deployment.
      • For more information on ReleaseOps, see ReleaseOps.
    • ServiceNow SDK:
      • Use the ServiceNow SDK to move applications to and from your instance to your local machine. You can integrate the ServiceNow SDK with your off-instance CI/CD process if you have one.
      • Install the ServiceNow SDK locally and use the command line interface (CLI).
      • Authenticate to a ServiceNow instance from the ServiceNow SDK.
      • Push to or install an application on the authenticated instance from your local environment.
    • Automated Test Framework (ATF)
      • Tests can be generated by Build Agent and executed in ServiceNow Studio or ServiceNow IDE to confirm functionality after changes.
      • For more information on ATF, see Automated Test Framework (ATF).