# Learn about Airship intelligence and AI features Explore AI-powered features for automated content generation, predictive send-time optimization, and analytics that surface which behaviors drive conversions. # AI in Airship > Airship uses AI to help you create better customer experiences faster. Our AI-powered features enhance efficiency and personalization while maintaining your brand identity and following responsible AI practices. ## Airship's AI principles Airship's AI capabilities are built on principles of responsible innovation that prioritize trust, transparency, and human oversight. Our approach ensures that while AI enhances your marketing efficiency and personalization capabilities, you maintain complete control over your brand voice and customer interactions. We implement robust security measures and strict data privacy protections, ensuring that customer data is never used to train foundational AI models. "Human in the Loop" principles are built into the platform by design. Our AI provides intelligent suggestions and starting points, but you always make the final decisions about content, targeting, and campaign execution. We've established comprehensive guardrails to ensure AI-generated content meets safety standards and avoids harmful material, while our continuous monitoring processes help maintain ethical AI practices. As global regulatory requirements develop, Airship remains committed to aligning our services with emerging compliance standards. This approach gives you confidence that our AI-powered solutions support business goals while upholding the highest ethical standards. For more information about our AI principles, see [AI-Powered Customer Experiences: Airship's Commitment to Responsible Innovation](https://www.airship.com/resources/whitepaper/ai-powered-customer-experiences-airships-commitment-to-responsible-innovation/). See also [Airship Security Measures](https://www.airship.com/legal/security-measures/). ## AI features AI features are available on [AXP](https://www.airship.com/docs/reference/feature-packages/) plans only. Airship offers the following features that use generative AI: * **Brand guidelines** — Ensure consistent brand identity by defining a profile, design elements, and personalities. Design elements automatically apply to specific message types, and you can also select them when creating messages. The profile and personalities are combined to create content for AI-generated Journeys. See [Set brand guidelines from uploaded sources](https://www.airship.com/docs/guides/messaging/features/brand-guidelines/#set-brand-guidelines-from-uploaded-sources) and [Set personalities](https://www.airship.com/docs/guides/messaging/features/brand-guidelines/#set-personalities) in *Setting brand guidelines*. * **Generative AI content** — AI-generated content is created from the text you provide. To refine the output, you can select a personality defined in your project's [Brand guidelines](https://www.airship.com/docs/guides/messaging/features/brand-guidelines/), include emojis :sunglasses:, and set the content length. The tool is available for most text fields when composing messages and templates. See [Generative AI content](https://www.airship.com/docs/guides/features/intelligence-ai/ai-content/). * **AI-generated Journeys** — A Journey is a continuous user experience of connected Sequences, Scenes and/or In-App Automations. You can use Generative AI to create draft Journeys. After generation, the map displays all linked components. See [Create AI-generated Journeys](https://www.airship.com/docs/guides/features/orchestration-experimentation/journeys/#create-ai-generated-journeys) in *Journeys*. We also provide AI agents that employ generative AI: * **Campaigns AI agent** — Create or refine a [Campaign](https://www.airship.com/docs/reference/glossary/#campaign) through a conversational chat interface. Generate a campaign overview with your objective, audience, and goals, and draft or update messages from a prompt or by uploading a file. See [Create and refine Campaigns using AI](https://www.airship.com/docs/guides/messaging/features/campaigns/#create-and-refine-campaigns-using-ai) in *Campaigns*. * **AI-generated Scene content** — Use the Native Experience AI Agent to generate and refine [Scene](https://www.airship.com/docs/reference/glossary/#scene) content through a conversational chat interface. Create screens from a prompt or mockup, set up [Branching](https://www.airship.com/docs/reference/glossary/#branching) paths, generate copy that matches your brand’s tone, and configure root appearance settings. See [Create Scene content using AI](https://www.airship.com/docs/guides/messaging/editors/native/ai-content/). * **AI Accessibility agent** — The accessibility agent uses agentic and generative AI to automatically audit your content and identify potential accessibility issues, such as insufficient color contrast, missing alternative text, and text size that falls below recommended minimums. Addressing these issues helps ensure your content is usable by the greatest number of audience members, including those with disabilities. See [Accessibility audit](https://www.airship.com/docs/guides/messaging/editors/native/about/#accessibility-audit) in *Native Experience editor*. * **AI recommendations** — Apply AI analysis to [Audience Pulse](https://www.airship.com/docs/reference/glossary/#audience_pulse) for deeper insights into audience activities and get actionable recommendations to build more effective, targeted campaigns. [AI recommendations](https://www.airship.com/docs/guides/audience/segmentation/audience-pulse/#ai-recommendations) in *Audience Pulse*. ## AI tools Airship provides two developer AI tools for working with the platform from AI coding assistants: * An MCP server gives assistants live access to Airship APIs and documentation. * Skills are pre-built workflows that guide assistants through common implementation tasks. For more information, see [AI tools for Airship developers](https://www.airship.com/docs/developer/ai-tools/ai-tools/). # Generative AI content > Use AI to create message content faster and according to guidelines you set. {{< badge "axp" >}} {{< badge "ai" "Generative AI" >}} AI-generated content is created from the text you provide. To refine the output, you can select a personality defined in your project's [Brand guidelines](https://www.airship.com/docs/guides/messaging/features/brand-guidelines/), include emojis :sunglasses:, and set the content length. The tool is available for most text fields when composing messages and templates. Use AI-generated text for quicker variant creation in A/B tests. See [About A/B testing](https://www.airship.com/docs/guides/experimentation/a-b-tests/about/). > **Note:** **Opting In to AI Functions** > > If you opted out of AI usage, you must sign an updated contract to enable this feature. Contact your account manager for assistance. > > **Compliance Considerations in Using AI Functions** > > The Service incorporates AI functions, including Generative AI and Agentic AI. > > Generative AI generates content such as Notification copy, images, and Journeys based on your prompts. > > Agentic AI autonomously optimizes, personalizes, or executes cross-channel customer engagement actions, or analyzes audience and performance data, subject to the parameters and controls you set in the Service. These systems operate under human-defined parameters and do not initiate customer-facing actions without human interaction or pre-configured parameters. You are responsible for reviewing Generated Outputs for accuracy, appropriateness, and to ensure they do not violate third-party intellectual property or other rights. Airship does not publish Generated Outputs to end users without approval from the Customer. > > In addition to the applicable terms of your agreement with Airship (e.g., Use of Service, Customer Responsibilities sections), you must comply with the [Airship Acceptable Use Policy](https://www.airship.com/legal/acceptable-use/), which provides additional details about appropriate conduct when using the Service. > > The Service includes safety features to block harmful content, such as content that violates our Acceptable Use Policy. You may not attempt to bypass these protective measures or use content that violates your agreement with Airship. > > About the AI models: > > Airship utilizes Google Gemini and Imagen to generate copy and images for AI Scene screens. The content is created solely with Google's out-of-the-box models, and no customization or fine-tuning with Customer Data is applied. See [Responsible AI](https://cloud.google.com/responsible-ai?hl=en) in Google's *Google Cloud* documentation. ## Generate AI content For each of your projects, you can generate AI content a maximum of 150 times per day. For a text field: 1. ![Accessing AI content generation](https://www.airship.com/docs/images/ai-content-icon_hu_f427f56f271f1e69.webp) *Accessing AI content generation* Select the AI content generation icon (✨) for the field. 1. ![Generating AI content](https://www.airship.com/docs/images/ai-content-form_hu_214779882537a227.webp) *Generating AI content* Under **AI Writing Assistant**, enter the information you want to include in your message. This can be a full version of the intended content or just keywords. The text you enter determines the language, but you can specify a different language for the generated content. Example text for a message: "A persuasive promotional email in Spanish encouraging users to join our new loyalty program and highlighting sign-up bonuses". 1. Make settings selections: | Setting | Description | | --- | --- | | **Include emoji** | Optional. Related emoji will be included in the generated text. | | **Brand personality** | The message content voice and tone will be determined by a personality defined in your [brand guidelines](https://www.airship.com/docs/guides/messaging/features/brand-guidelines/). If none is selected, voice and tone will be determined only from your provided text. | | **Content length** | Your message can be Short (~80 characters), Medium (~240 characters), or Long (~400 characters). | 1. Select **✨ Generate content** and review the generated version. If it's not quite right, select **Try again**, edit your provided text, adjust the settings, and then generate content again. 1. Select **Use this text** to transfer the current generated content to the message's text field, where you can refine your text. > **Note:** Review generated content transferred to text fields before sending a message. ## Predictive Analytics Use predictive scores for churn risk and optimal send time to improve campaign performance and audience targeting. # Predictive Churn > {{< glossary_definition "predicted_to_churn" >}} Churn is a natural part of engagement ebb and flow, and while a certain amount of churn is normal and healthy, there are ways to identify churn risk factors and take actions to prevent your user base from eroding. ## About Predictive Churn With Predictive Churn, you can identify app and web users by their likelihood to churn, based on risk profiles Airship generates via machine learning, using [gradient boosted decision tree](https://en.wikipedia.org/wiki/Gradient_boosting) methodology. Risk factors update weekly and are exposed as [Tags](https://www.airship.com/docs/reference/glossary/#tag) for [segmentation](https://www.airship.com/docs/guides/audience/segmentation/segmentation/) and analysis in [Performance Analytics](https://www.airship.com/docs/reference/glossary/#pa), and exposed as Tag Change events in [Real-Time Data Streaming](https://www.airship.com/docs/reference/glossary/#rtds). Example use cases for Predictive Churn: * Target users with offers before they churn. * Run an A/B Test with a single variant and a control group to measure the message's impact on churn. * Trigger an automation or [Sequence](https://www.airship.com/docs/reference/glossary/#sequence) based on a change in risk group. * Send a message in a Sequence based on a change in risk group. * Create a [Segment](https://www.airship.com/docs/reference/glossary/#segment) that blends risk groups based on the type of messaging and your goals. ### The Predictive Churn model Predictive Churn belongs to Airship's *Predictive* suite of products, which uses [machine learning](https://en.wikipedia.org/wiki/Machine_learning) to predict user behaviors and optimize engagement strategy for customer lifecycle marketers. The model is trained on recency and frequency of notification sends and app opens or website visits for a cross-section of anonymized apps and sites. It runs weekly on Mondays. It detects the most relevant risk factors for a churn outcome and assigns a high, medium, or low churn factor to each user who has been active in the past 60 days. By including your app key as an input, the model tailors its predictions to your audience based on your app or website usage. Notable terms: Active User : An *active user* is a member of your audience that has opened your app, had an active web session, or clicked a web notification in the last 30 days. Inactive User : An *inactive user* is a member of your audience that had a predictive tag of high, medium, or low and has not opened your app, had an active web session, or clicked a web notification in the last 30 days. Churn : A *churn* outcome occurs when a previously active user becomes inactive, i.e., Airship has not seen any activity (measured in app opens, website visits, or web notification clicks) from a user in the last 30 days. Push opt-in status does not factor into the churn outcome, so it is possible that a user who opted out of notifications could still appear active for churn prediction purposes. > **Note:** A churned user is not the same as an uninstalled user. Churn Risk : Predictive Churn makes a prediction about the likelihood of a future churn outcome, meaning that a user will go inactive. We assign one of three measures of risk for such an outcome: * High — Users most likely to become inactive * Medium — Users who exhibit signs of potentially becoming inactive * Low — Users least likely to become inactive ### Tags and change events A user's churn risk profile is represented as a `high`, `medium`, or `low` [Tag](https://www.airship.com/docs/reference/glossary/#tag) within the `ua_churn_prediction` [Tag Group](https://www.airship.com/docs/reference/glossary/#tag_group). Changes to that tag are represented as `TAG_CHANGE` events. Tag changes return both the change in tag (`add` or `remove`) and the `current` tag. The `current` tag is the end result of the tag change. There are three scenarios for tag change events: * **Add prediction** — Adds a new Predictive Churn tag to a channel that did not previously have a prediction. Not all devices begin with a churn prediction; churn prediction is assigned to active users when the Predictive Churn model runs (weekly on Mondays). * **Prediction change** — Replaces the prediction on a channel. * **Remove prediction** — Removes the prediction from a channel, typically when a channel becomes inactive. The following is an example of a Predictive Churn tag change: ```json { "id": "e1559cd7-af96-45ab-bb74-a22cd99a01d5", "offset": "1422600", "occurred": "2017-01-15T09:26:30.362Z", "processed": "2017-01-15T16:15:30.048Z", "device": { "android_channel": "d5ec96e3-5ced-47b0-a4dd-1b2b6bda442e", "named_user_id": "job.bob@example.com", "attributes": { "locale_variant": "", "app_version": "312", "device_model": "LG-H811", "app_package_name": "com.company.app", "iana_timezone": "America/Los_Angeles", "push_opt_in": "true", "locale_country_code": "US", "device_os": "6.0", "locale_timezone": "-28800", "locale_language_code": "en", "location_enabled": "true", "background_push_enabled": "true", "ua_sdk_version": "6.1.2", "location_permission": "ALWAYS_ALLOWED" } }, "body": { "add": { "ua_churn_prediction": [ "medium" ] }, "remove": { "ua_churn_prediction": [ "high" ] }, "current": { "ua_churn_prediction": [ "medium" ] } }, "type": "TAG_CHANGE" } ``` See the [Tag change event](https://www.airship.com/docs/developer/rest-api/connect/schemas/events/#tag-change) in the [Real-Time Data Streaming](https://www.airship.com/docs/reference/glossary/#rtds) API reference. ## Use Predictive Churn in messaging Before you can use Predictive Churn for targeting, you must enable it for your project. It is supported for production projects only. If your app and website both use the Airship SDK, you should enable the feature for both. 1. Next to your project name, select the dropdown menu (▼), then **Settings**. 1. Under **Project settings**, select **Predictive AI**. 1. Enable **Predictive App Churn** and/or **Predictive Web Churn**. Tags are assigned the first Monday after enabling the feature. ### Segments You can include a Predictive Churn risk profile in your [Segments](https://www.airship.com/docs/reference/glossary/#segment). First, search for and select **Predicted to Churn**, and then select an operator and a risk profile. ### Message and experiment audiences Add a Predictive Churn risk profile as an audience condition for messages and experiments. See **Predicted to Churn** in [Targeting Specific Users](https://www.airship.com/docs/guides/audience/segmentation/target-specific-users/). You can also specify a risk profile for individual Sequence messages. See **Conditions** in [Add messages to a Sequence](https://www.airship.com/docs/guides/messaging/messages/sequences/create/add-messages/). ### Automation and Sequence triggers Trigger an [Automation](https://www.airship.com/docs/guides/messaging/messages/sequences/create-automation/) or [Sequence](https://www.airship.com/docs/reference/glossary/#sequence) based on changes to a user's Predictive Churn risk profile. For example, you might set up an automation to send users a special offer when their Predictive Churn risk changes to High, helping retain users at risk of leaving your audience. See **Predicted to Churn** in [Automation and Sequence triggers](https://www.airship.com/docs/guides/messaging/messages/sequences/triggers/). For automation using the API, see the next section. You can also set a churn risk profile as a trigger condition. For Automations, see **Conditions** in [Create an Automation](https://www.airship.com/docs/guides/messaging/messages/sequences/create-automation/). For Sequences, see **Conditions** in the Trigger step in [Create a Sequence](https://www.airship.com/docs/guides/messaging/messages/sequences/create/create/). ### Audience definition in the API With the API, use the [audience tag selector object](https://www.airship.com/docs/developer/rest-api/ua/schemas/audience-selection/) to target by the `ua_churn_prediction` Tag Group and Tag values `low`, `medium`, or `high`. For example, the following is a notification to users of all device types whose current churn prediction is `medium`: ```http POST /api/push HTTP/1.1 Authorization: Basic Content-Type: application/json Accept: application/vnd.urbanairship+json; version=3 { "audience": { "tag": "medium", "group": "ua_churn_prediction" }, "notification": { "alert": "me·di·um, n., an agency or means of doing something." }, "device_types": [ "ios", "android", "web" ] } ``` ## Analytics In [Performance Analytics](https://www.airship.com/docs/reference/glossary/#pa), the [Predictive Dashboard](https://www.airship.com/docs/guides/reports/analytics/definitions/#predictive) helps you track churn risk factors over time. This Dashboard provides a view into Predictive Churn risk groups, distribution of users across risk groups, and the performance of churn mitigation tactics. If you have both Predictive App and Web Churn enabled, you can set the Device Family filter to Web or Mobile to see churn data for either audience. Predictive tags update every Sunday, and reports default to the most recent update. **Use cases:** * Explore added or removed Predictive tags. * Slice user behavior by churn risk tag. * Export ad IDs, named users, and channel IDs based on their risk category. * Export named users and ad IDs based on app opens, uninstalls, and risk category. * Find churn cohorts and slice by the users' current tags. * Find churn cohorts, filter, then analyze a funnel of past behavior. # Optimal Send Time > Optimal Send Time is an algorithm that determines the best hour for optimal engagement activity — when each individual member of your audience is most likely to receive and act on your message. Take the guesswork out of scheduling messages and let Airship's predictive models optimize send times for you. Send time predictions update weekly and are exposed as [Tags](https://www.airship.com/docs/reference/glossary/#tag) for [segmentation](https://www.airship.com/docs/guides/audience/segmentation/segmentation/) and analysis in [Performance Analytics](https://www.airship.com/docs/reference/glossary/#pa), and exposed as Tag Change events in [Real-Time Data Streaming](https://www.airship.com/docs/reference/glossary/#rtds). > **Note:** Optimal Send Time is available for iOS, Android, and Fire OS only. ## The Optimal Send Time model Optimal send time is determined from recent engagement history. To start, app opens are localized to the user's time zone and aggregated to the hour over the last 60 days of app activity. The best hour is determined by striking a balance between the user's engagement patterns and a generalized model of engagement patterns across the app audience. The model also outputs a general best hour, which is applied to dormant or low-activity users. The general best hour aggregates opens across app users and selects the best hour based on more frequent opening time for each app platform. The determined best hour will occur within a three-day window around your selected delivery date. For example, if you choose February 2, the message will be delivered between February 1 and 3. After enabling the feature, Airship runs the predictive model for your iOS, Android, and Fire OS audience members, and you can start using the Optimal Send Time model for delivery. ## Optimal Send Time use cases Schedule notifications without having to guess the optimal time for user engagement. By delivering a message to your users at the best time for them, you can optimize for a higher open rate. * Send an important update to all users at the time they are most likely to read your message. * Deliver a coupon to your users at a time when they are most likely to engage. * Send a long form story to you readers at the best time for them. * Distribute user engagement across the day to meter traffic flow to the app. * Compare performance between regular scheduled messages and messages sent using Optimal Send Time. * Analyze Optimal Send Time user level distribution across hours of the day. * Analyze correlation between churn risk and user’s best send time. ## Optimal Send Time data and analytics The Optimal Send Time dashboard in Performance Analytics provides a deeper look into the best time model, including a distribution of best hours across your audience, and the generalized best hour for your audience by platform and day of week. You can also use Real-Time Data Streaming to observe changes in optimal send time as `TAG_CHANGE` events for the `ua_send_time_prediction` tag group. ## Enable Optimal Send Time

First you must enable Optimal Send Time in the dashboard. Send time prediction supports your production projects only and updates weekly on Wednesdays.

  1. Next to your project name, select the dropdown menu (▼), then Settings.
  2. Under Project settings, select Predictive AI.
  3. Enable Optimal Send Time.
## Schedule a message using Optimal Send Time

You can use Optimal Send Time in the Message and A/B Test composers. In the Delivery step:

  1. Select Optimize and enter a date.

OR (Message composer only)

  1. Select Recurring.
  2. Specify the delivery interval by number of hours/days/weeks/months/years.
  3. Set the date for the initial delivery. This is the first time that Airship will send your message.
  4. Select Optimal time.

Airship recommends scheduling your message at least three days in advance due to the combination of time zones and optimal times. You can reduce the lead time if your audience is more localized, e.g., only in the United States or in a certain European region.

> **Note:** When your audience includes users without an optimal send time tag, those users will be dropped from delivery and will not receive the message. Since optimal send time is determined from user behavior over time, new users might not have an optimal send time determined for the first week or two after channel registration. ## Schedule `best_time` messages using the API

In the API, Optimal Send Time is represented as the best time key. To deliver notifications at your users’ optimal times via the API, schedule your message using best_time.

The following example shows two schedules for an upcoming message. The first schedule uses a specific scheduled_time for users with the “earlyBirds” tag, and the second schedule lets the model decide when to send the message, based on the best_time for users with the “normalPeople” tag.

```http POST /api/schedules HTTP/1.1 Authorization: Basic Content-Type: application/json Accept: application/vnd.urbanairship+json; version=3 [ { "name": "Morning People", "schedule": { "scheduled_time": "2018-06-03T09:15:00" }, "push": { "audience": { "tag": "earlyBirds" }, "notification": { "alert": "Good Day Sunshine" }, "device_types": [ "ios", "android", "sms", "web" ] } }, { "name": "Everybody Else", "schedule": { "best_time": { "send_date": "2018-06-03" } }, "push": { "audience": { "tag": "normalPeople" }, "notification": { "alert": "Stay Up Late" }, "device_types": [ "ios", "android", "sms", "web" ] } } ] ``` ## Look up a user's Optimal Send Time You can look up optimal send time for an individual user from the dashboard. See also: [Contact management](https://www.airship.com/docs/guides/audience/contact-management/). 1. Go to *Audience » Contact Management*. 1. Enter a [Channel ID](https://www.airship.com/docs/reference/glossary/#channel_id), [Named User ID](https://www.airship.com/docs/reference/glossary/#named_user), or [Device Token](https://www.airship.com/docs/reference/glossary/#device_token). 1. Click a result to view the named user ID or channel ID page, then click the channel ID and go to the *Tag Groups* tab. The hour tag is listed in the UA_SEND_TIME_PREDICTION tag group. In the API, the `ua_send_time_prediction` tag group represents Optimal Send Time. You can look up a channel or named user and check the `ua_send_time_prediction` tag group to find the optimal send time for a user. See [Channel Lookup](https://www.airship.com/docs/developer/rest-api/ua/operations/channels/#getchannel) and [Named User Listing or Lookup](https://www.airship.com/docs/developer/rest-api/ua/operations/named-users/#getnameduser) in our API documentation. ## Observe Optimal Send Time in Performance Analytics The Optimal Send Time dashboard in Performance Analytics provides a deeper look into the best time model, including a distribution of best hours across your audience, and the generalized best hour for your audience by platform and day of week. Go to the Optimal Send Time dashboard: 1. Go to *Reports » Performance Analytics*. 1. Go to *Spaces » Shared » Predictive* and select *Optimal Send Time*. ## Optimal Send Time events The `ua_send_time_prediction` tag group contains the send time prediction for each channel. Changes in a user's `ua_send_time_prediction` tag appear as Tag Change Events in the event stream.