Deploy

Discover how to create a build and deploy your conversational application in your NLX workspace

What's the Deployment tab of an application?

An application's Deployment tab allows you to construct a build of your conversational AI application that contains all the current versions of the flows attached, training data, channel set up, language data and settings, and NLP settings. You may then deploy a successful build to production or roll back to a previous deployment.

Your chosen channels control how applications are rendered or experienced by users, as they provide the frontend interface. Deployments communicate any changes externally to channels.

A Natural Language Processing (NLP) engine must be connected to your workspace before engaging with the Deployment tab. See how to integrate NLP.

Create a build

Builds are packaged versions of your application that include all resources associated in the state they look like at the time the build is created:

To create an application build:

  • Click Deployment tab of application > Select Review & build

  • Review Validation check for critical errors or detected UX issues

  • Select and ensure NLP engine is correct

Optional

  • Enable NLX Boost: Enhances the performance of your NLP engine with generative AI intent classification

    • Allow NLX Boost to override the NLP: Enabling this relies on NLX's built-in generative AI to detect user intent (based on your training data) and route to a flow accordingly, regardless of NLP's detected match

  • Click Create build

You may now deploy your build at any time to make it live to users.

Validation check

Before each new build initiates, a validation check is run to provide a preview of potential errors that may cause failed builds or issues that may result in broken flows or poor UX. Detected issues are listed with descriptions and potential solutions:

Build status

The build status column indicates whether a build was successful or failed. Only successful builds may be deployed:

🟢 Successful build: Available to deploy, making your application available outside of NLX on configured channel(s)

🔴 Failed build: Unavailable to deploy. Select Details next to the build status to expand information on what caused the failure.

Once a build is made, resources (Flows, Data requests, etc.) can be edited without affecting that particular build. You may also access a previous build to make available to users.

Deploy a build

For each successful build, a Deploy option is available to make that build live outside of NLX, effectively making your application live and in production:

  • Click Deploy

Touchpoint configurator or MCP client setup

Only one build can be deployed at a time and deploying any build deactivates the previous one. You can freely alternate between newer and older builds using Rollback or Deploy.

Test a build

When a build is successful, you can engage your application in a test conversation using the Test feature in NLX. The Test chat only runs through the latest build of your application, as each new build updates the NLP engine you've selected. Rollbacks to an earlier build are not reflected in Test simulations.

For Google Dialogflow applications, be sure to create an agent and deploy it in your Dialogflow console, then paste the agent's Google project ID into the application's languages tab in NLX.

For more details, see Test

Delete a deployment

To take your application offline, you may delete the current deployment:

  • Select Details next to the the deployed build

  • Under Danger zone, click Delete deployment

Until you redeploy, your application remains offline.

Last updated