Andy in the Cloud

From BBC Basic to Force.com and beyond…


10 Comments

Getting your users attention with Custom Notifications

customnotificationsummaryGetting your users attention is not always easy, choosing how, when and where to notify them is critical. Ever since Lightning Experience and Salesforce Mobile came out the notification bell has been a one stop shop for Chatter and Approval notifications, regardless if you are on your desktop or your mobile device.

In beta release at time of writing is a new platform feature known as Notification Manager that allows you to send your own custom notifications to your users for anything your heart desires from the very same locations, even on a users mobile device! This blog dives into this feature and how you can integrate it into your creations regardless if you are a admin click coder, Apex developer or REST API junkie.

Getting Started

The first thing you need to do is define a new Notification Type under the Setup menu. This is a simple process that involves giving it a name and deciding what channels you want the notification to go out on, currently user desktop and mobile devices.

notificationstypes

Once this has been done you can use the new Send Custom Notification action in Process Builder or Flow. This allows you to define the title and body of your notification, along with the target recipients (users, groups, queues and more) along the target record that determines the record the user sees when they click/tap the notification. The following screenshot shows an example of such an Action in Process Builder:-

opportunitycustomnotification

notificationoppty.png

Basically that is all there is to it! You will have in a few clicks empowered yourself with the ability to reach out to not only your users desktop but the actual mobile device notification experience on each of the their mobile devices!  You didn’t have to learn how to write a mobile app, figure out how to do mobile notifications, register things with Google or Apple. I am honestly blown away at how easy and powerful this is!

So it is pretty easy to to send notifications this way from Process Builder processes driven by record updates from the user and also reference field values to customize the notification text. However in the ever expanding world of Platform Events, how do we send custom notifications based on Platform Events?

Sending Custom Notifications for Batch Apex Job Failures

One of my oldest and most popular blog posts discussed design best practices around Batch Apex jobs. One of the considerations it calls out is how important it is to route errors that occur in the background back to the user. Fast forward a bit to this blog, where I covered the new BatchApexError Platform Event as a means to capture and route batch errors (even uncatchable exceptions) in near realtime. It also describes strategy to enabled users to retry failed jobs. What it didn’t really solve, is letting them know something had gone wrong without them checking a custom tab. Let’s change that!

Process Builder is now able to subscribe to the standard BatchApexErrorEvent and thus enables you as an admin to apply filter and routing logic on failed batch jobs. When combined with custom notifications those errors can now be routed to users devices and/or desktops in realtime. While Process Builder can subscribe to events it does have some restrictions on what it can do with the event data itself. Thus we are going to call an autolaunch Flow from Process Builder to actually handle the event and send the custom notification from within Flow. If you are reading this wondering if your Apex code can get in on the action, the answer is yes (ish), more on this later though. The declarative solution utilizes one Process Builder process and two Flows. The separation of concerns between them is shown in the diagram below:-

batchapexeventtonotificationarch

Let’s work from the bottom to the top to understand why I decided to split it up this way. Firstly, SendCustomNotification is a Sub Flow (callable by other Flows) and is a pretty simple wrapper around the new Send Custom Notification action shown above. It does help with one wrinkle when working within Flow. You have to pass a Notification Type Id. In this case this Sub Flow leverages a small Custom Metadata Type map that contains the predefined Notification Names and Ids. This magic is encapsulated in the Sub Flow, so if that restriction is lifted in the future, this can easily be removed. You can take a closer look at this later through the sample code repository here.

SendCustomNotification

Next the BatchApexErrorPlatformEventHandler Flow defines a set of input variables that are populated from the Process Builder process. These variables match the fields and types per the definition of the Batch Apex Error Event here. The only other thing it does is add the Id of the user that generated the event (aka the user who submitted the failed job) to the list of recipients passed to the SendCustomNotification sub flow above. This could also be a Group Id if you wanted to send the notification further.

batchapexeventhandlerflow.png

Lastly, in the screenshot below you see the Process Builder that subscribes to the Batch Apex Error Event and maps the event field values to the input variables exposed from BatchApexErrorPlatformEventHandler Flow via the EventReference. The example here is very simple, but you can now imagine how you can add other filter criteria to this process that allows you to inspect which Batch Apex job failed and route and/or adjust messaging in the notifications accordingly, all done declaratively of course!

batchapexerrorfrompb.png

NOTE: It is not immediately apparent in all cases that you can access the event fields from Process Builder, since the documentation states them as not supported within formulas. I want to give a shout out to Alex Edelstein PM for Flow for clarifying that it is possible! Check out his amazing blog around all things Flow here. Finally note that Process Builder requires an Object to map the incoming event to. In this case I mapped to a User record using the CreatedById field on the event.

Sending Custom Notifications from Code

The Send Custom Notification action is also exposed via the Salesforce Action REST API defined here (hint hint for Doug Ayers Mass Action tool to support it). You can of course attempt to call this REST API via Apex as well. While there is currently no native Apex Action API, it turns out calling the above SendCustomNotification Flow from Apex works pretty well meanwhile. I have written a small wrapper around this technique to make it a little more elegant to perform from Apex and it also serves to abstract away this hopefully temporary workaround for Apex developers.

new CustomNotification()
    .type('MyNotificationType')
    .title('Fun Custom Notification')
    .body('Custom Notifications are Awesome!')
    .sendToCurrentUser();

The following Apex code results in this notification appearing on your device!

funfromapexnotification.png

This CustomNotification helper class is included in the sample code for this blog and leverages another class I wrote that wraps the native Apex Flow API. I used this wrapper because it allowed me to mock the actual Flow invocation since there is no way as far as I can see to assert the notification was actually sent.

NOTE: When sending custom notifications via declarative tools and/or via code I did confirm in my testing that they are included in the current transaction. Also I recommend you always avoid calling Flow in loops in your Apex code, instead make your Flows take list variables (aka try to bulkify Flows called from Apex). Though not shown in the Apex above, the wrapped Flow takes a list of recipients.

Summary

So there you have it, custom mobile and desktop notifications sent from Process Builder, Flow, Apex and REST API. Keep in mind of course at time of writing this is a Beta feature and thus read the clause in the documentation carefully. Now go forth and start thinking of all the areas you can enable with this feature!

P.S. Check out another new cool feature called Lightning In-App Guidance.

 

 


5 Comments

Adding Clicks not Code Extensibility to your Apex with Lightning Flow

Building solutions on the Lightning Platform is a highly collaborative process, due to its unique ability to allow Trailblazers in a team to operate in no code, low code and/or code environments. Lightning Flow is a Salesforce native tool for no code automation and Apex is the native programming language of the platform — the code!

A flow author is able to create no-code solutions using the Cloud Flow Designer tool that can query and manipulate records, post Chatter posts, manage approvals, and even make external callouts. Conversely using Salesforce DX, the Apex developer can, of course, do all these things and more! This blog post presents a way in which two Trailblazers (Meaning a flow author and an Apex developer) can consider options that allow them to share the work in both building and maintaining a solution.

Often a flow is considered the start of a process — typically and traditionally a UI wizard or more latterly, something that is triggered when a record is updated (via Process Builder). We also know that via invocable methods, flows and processes can call Apex. What you might not know is that the reverse is also true! Just because you have decided to build a process via Apex, you can still leverage flows within that Apex code. Such flows are known as autolaunched flows, as they have no UI.

Blog_Graphic_01_v01-02_abpx5x.png

I am honored to have this blog hosted on the Salesforce Blog site.  To continue reading the rest of this blog head on over to Salesforce.com blog post here.

 


19 Comments

Simplified API Integrations with External Services

Salesforce are on a mission to make accessing off platform data and web services as easy as possible. This helps keep the user experience optimal and consistent for the user and also allows admins to continue to leverage the platforms tools such as Process Builder and Flow, even if the data or logic is not on the platform.

Starting with External Objects, they added the ability to see and also update data stored in external databases. Once setup, users can manipulate external records without leaving Salesforce, by staying within the familiar UI’s. With External Services, currently in Beta, they have extended this concept to external API services.

UPDATE: The ASCIIArt Service covered in this blog has since been updated to use the Swagger schema standard. However this blog is still a very useful introduction to External Services. Once you have read it, head on over to this blog!

In this blog lets first focus on the clicks-not-code steps you can repeat in your own org, to consume a live ASCII Art web service API i have exposed publicly. The API is simple, it takes a message and returns it in ASCII art format. The following steps result in a working UI to call the API and update a record.

ExternalServicesDemo.png

After the clicks not code bit i will share how the API was built, whats required for compatibility with this feature and how insanely easy it is to develop Web Services in Heroku using Nodejs. So lets dive in to External Services!

Building an ASCII Art Converter in Lightning Experience and Flow

The above solution was built with the following configurations / components. All of which are accessible under the LEX Setup menu (required for External Services) and takes around 5 minutes maximum to get up and running.

  1. Named Credential for the URL of the Web Service
  2. External Service for the URL, referencing the Named Credential
  3. Visual Flow to present a UI, call the External Service and update a record
  4. Lightning Record Page customisation to embed the Flow in the UI

I created myself a Custom Object, called Message, but you can easily adapt the following to any object you want, you just need a Rich Text field to store the result in. The only other thing you need to know of course is the web service URL.

https://createasciiart.herokuapp.com

Can i use External Services with any Web Service then?

In order to build technologies that simplify what are normally things developers have to interpret and code manually. Web Service APIs must be documented in a way that External Services can understand. In this Beta release this is the Interagent schema standard (created by Heroku as it happens).  Support for the more broadly adopted Swagger / OpenId will be added in the Winter release (Safe Harbour).

For my ASCII Art service above, i authored the Interagent schema based on a sample the Salesforce PM for this feature kindly shared, more on this later. When creating the External Service in moment we will provide a schema to this service.

https://createasciiart.herokuapp.com/schema

Creating a Named Credential

From the setup menu search for Named Credential and click New. This is a simple Web Service that requires no authentication. Basically provide only the part of the above URL that points to the Web Service endpoint.

ESNamedCred.png

Creating the External Service

Now for the magic! Under the Setup menu (only in Lightning Experience) search for Integrations and start the wizard. Its a pretty straight forward process, of selecting the above Named Credential, then telling it the URL for the schema. If thats not exposed by the service you want to use, you can paste a Schema in directly (which lets a developer define a schema yourself if one does not already exist).

esstep1.png

esstep2.png

Once you have created the External Service you can review the operations it has discovered. Salesforce uses the documentation embedded in the given schema to display a rather pleasing summary actually.

esstep3.png

So what just happened? Well… internally the wizard wrote some Apex code on your behalf and implemented the Invocable Method annotations to enable that Apex code to appear in tools like Process Builder (not supported in Beta) and Flow. Pretty cool!

Whats more interesting for those wondering, is you cannot actually see this Apex code, its there but some how magically managed by the platform. Though i’ve not confirmed, i would assume it does not require code coverage.

Update: According to the PM, in Winter’18 it will be possible “see” the generated class from other Apex classes and thus reuse the generated code from Apex as well. Kind of like a Api Stub Generator.

Creating a UI to call the External Service via Flow

This simple Flow prompts the user for a message to convert, calls the External Service and updates a Rich Text field on the record with the response. You will see in the Flow sidebar the generated Apex class generated by the External Service appears.

esflow

The following screenshots show some of the key steps involved in setting up the Flow and its three steps, including making a Flow variable for the record Id. This is later used when embedding the Flow in Lightning Experience in the next step.

esflow1

RecordId used by Flow Lightning Component

esflow2

Assign the message service parameter

esflow3

Assign the response to variable

esflow4

Update the Rich Text field

TIP: When setting the ASCII Art service response into the field, i wrapped the value in the HTML elements, pre and code to ensure the use of a monospaced font when the Rich Text field displayed the value.

Embedding the Flow UI in Lightning Experience

Navigate to your desired objects record detail page and select Edit Page from the cog in the top right of the page to open the Lightning App Builder. Here you can drag the Flow component onto the page and configure it to call the above flow. Make sure to map the Flow variable for the record Id as shown in the screenshot, to ensure the current record is passed.

esflowlc.png

Thats it, your done! Enjoy your ASCII Art messages!

Creating your own API for use with External Services

Belinda, the PM for this feature was also kind enough to share the sample code for the example shown in TrailheaDX, from which the service in this blog is based. However i did wanted to build my own version to do something different from the credit example. Also extend my personal experience with Heroku and Nodejs more.

The NodeJS code for this solution is only 41 lines long. It runs up a web server (using the very easy to use hapi library), and registers a couple of handlers. One handler returns the statically defined schema.json file, the other implements the service itself. As side note, the joi library is an easy way add validation to the service parameters.

var Hapi = require('hapi');
var joi = require('joi');
var figlet = require('figlet');

// initialize http listener on a default port
var server = new Hapi.Server();
server.connection({ port: process.env.PORT || 3000 });

// establish route for serving up schema.json
server.route({
  method: 'GET',
  path: '/schema',
  handler: function(request, reply) {
    reply(require('./schema'));
  }
});

// establish route for the /asciiart resource, including some light validation
server.route({
  method: 'POST',
  path: '/asciiart',
  config: {
    validate: {
      payload: {
        message: joi.string().required()
      }
    }
  },
  handler: function(request, reply) {
    // Call figlet to generate the ASCII Art and return it!
    const msg = request.payload.message;
    figlet(msg, function(err, data) {
        reply(data);
    });
  }
});

// start the server
server.start(function() {
  console.log('Server started on ' + server.info.uri);
});

I decided i wanted to explore the diversity of whats available in the Nodejs space, through npm. To keep things light i chose to have a bit of fun and quickly found an ASCIIArt library, called figlet. Though i soon discovered that npm had a library for pretty much every other use case i came up with!

Finally the hand written Interagent schema is also shown below and is reasonably short and easy to understand for this example. Its not all that well documented in layman’s terms as far as i can see. See my thoughts on this and upcoming Swagger support below.

{
  "$schema": "http://interagent.github.io/interagent-hyper-schema",
  "title": "ASCII Art Service",
  "description": "External service example from AndyInTheCloud",
  "properties": {
    "asciiart": {
      "$ref": "#/definitions/asciiart"
    }
  },
  "definitions": {
    "asciiart": {
      "title": "ASCII Art Service",
      "description": "Returns the ASCII Art for the given message.",
      "type": [ "object" ],
      "properties": {
        "message": {
          "$ref": "#/definitions/asciiart/definitions/message"
        },
        "art": {
          "$ref": "#/definitions/asciiart/definitions/art"
        }
      },
      "definitions": {
        "message": {
          "description": "The message.",
          "example": "Hello World",
          "type": [ "string" ]
        },
        "art": {
          "description": "The ASCII Art.",
          "example": "",
          "type": [ "string" ]
        }
      },
      "links": [
        {
          "title": "AsciiArt",
          "description": "Converts the given message to ASCII Art.",
          "href": "/asciiart",
          "method": "POST",
          "schema": {
            "type": [ "object" ],
            "description": "Specifies input parameters to calculate payment term",
            "properties": {
              "message": {
                "$ref": "#/definitions/asciiart/definitions/message"
              }
            },
            "required": [ "message" ]
          },
          "targetSchema": {
            "$ref": "#/definitions/asciiart/definitions/art"
          }
        }
      ]
    }
  }
}

Finally here is the package.json file that brings the whole node app together!

{
  "name": "asciiartservice",
  "version": "1.0.0",
  "main": "server.js",
  "dependencies": {
    "figlet": "^1.2.0",
    "hapi": "~8.4.0",
    "joi": "^6.1.1"
  }
}

Other Observations and Thoughts…

  • Error Handling.
    You can handle errors from the service in the usual way by using the Fault path from the element. The error shown is not all that pretty, but then in fairness there is not really much of a standard to follow here.
    eserrorflow.pngeserror.png
  • Can a Web Service called this way talk back to Salesforce?
    Flow provides various system variables, one of which is the Session Id. Thus you could pass this as an argument to your Web Service. Be careful though as the running user may not have Salesforce API access and this will be a UI session and thus will be short lived. Thus you may want to explore another means to obtain an renewable oAuth token for more advanced uses.
  • Web Service Callbacks.
    Currently in the Beta the Flow is blocked until the Web Service returns, so its good practice to make your service short and sweet. Salesforce are planning async support as part of the roadmap however.
  • Complex Parameters.
    Its unclear at this stage how complex a web service can be supported given Flows limitations around Invocable Methods which this feature depends on.
  • The future is bright with Swagger support!
    I am really glad Salesforce are adding support for Swagger/OpenID, as i really struggled to find good examples and tutorials around Interagent. Really what is needed here is for the schema and code to be tied more closely together, like this!UPDATE: See my other blog entry covering Swagger support

Summary

Both External Objects and External Services reflect the reality of the continued need for integration tools and making this process simpler and thus cheaper. Separate services and data repositories are for now here to stay. I’m really pleased to see Salesforce doing what it does best, making complex things easier for the masses. Or as Einstein would say…Everything should be made as simple as possible, but no simpler.

Finally you can read more about External Objects here and here through Agustina’s and laterally Alba’s excellent blogs.


2 Comments

Highlights from TrailheaDX 2017

IMG_2857.JPGThis was my first TrailheaDX and what an event it was! With my Field Guide in hand i set out into the wilderness! In this blog i’ll share some of my highlights, thoughts and links to the latest resources. Many of the newly announced things you can actually get your hands on now which is amazing!

Overall the event felt well organized, if a little frantic at times. With smaller sessions of 30 minutes each, often 20 mins after intros and questions, each was bite sized, but quite well tuned with demos and code samples being shown.

SalesforceDX, Salesforce announced the public beta of this new technology aimed at improving the developer experience on the platform. SalesforceDX consist of several modules that will be extended further over time. Salesforce has done a great job at providing a wealth of Trailhead resources to get you started.

Einstein, Since its announcement, myself and other developers have been waiting to get access to more custom tools and API’s, well now that wait is well and truly over. As per my previous blogs we’ve had the Einstein Vision API for a little while now. Announced at the event where no less than three other new Einstein related tools and API’s.

  • Einstein Discovery. Salesforce demonstrated a very slick looking tool that allows you to point and click your way through to analyzing various data sets, including those obtained from your custom objects! They have provided a Trailhead module on it here and i plan on digging in! Pricing and further info is here.
  • Einstein Sentiment API. Allows you to interpret text strings for terms that indicate if its a positive, neutral or negative statement / sentiment. This can be used to scan case comments, forum posts, feedback, twitter posts etc in an automated way and respond or be alerted accordingly to what is being said.
  • Einstein Intent API.  Allows you to interpret text strings for meanings, such as instructions or requests. Routing case comments or even implementing bots that can help automate or propose actions to be taken without human interpretation.
  • Einstein Object Detection API. Is an extension of the Einstein Vision API, that allows for individual items in a picture to be identified. For example a pile of items on a coffee table, such as a mug, magazine, laptop or pot plant! Each can then be recognized and classified to build up more intel on whats in the picture for further processing and analysis.
  • PredictionIO on Heroku. Finally, if you want to go below the declarative tools or intentional simplified Einstein API’s, you can build your own machine learning models using Heroku and the PredictionIO build pack!

Platform Events. These allow messages to be published and subscribed to using a new object known as an Event object, suffixed with __e. Once created you can use a new Apex API’s or REST API’s to send messages and use either Apex Triggers or Streaming API to receive them. There is also a new Process Builder action or Flow element to send messages. You can keep your messages within Force.com or use them to integrate between other cloud platforms or the browser. The possibilities are quite endless here, aysnc processing, inter app comms, logging, ui notifications…. i’m sure myself and other bloggers will be exploring them in the coming months!

External Services. If you find a cool API you want to consume in Force.com you currently have to write some code. No longer! If you have a schema that describes that API you use the External Services wizard to generate some Apex code that will call out to the API. Whats special about this, is the Apex code is accessible via Process Builder and Flow. Making clicks not code API integration possible. Its an excellent way to integrate with complementary services you or others might develop on platforms such as Heroku. I have just submitted a session to Dreamforce 2017 to explore this further, fingers crossed it gets selected! You can read more about it here in the meantime.

Sadly i did have to make a key decision to focus on the above topics and not so much on Lightning. I heard from other attendees these sessions where excellent and i did catch a brief sight of dynamic component rendering in Lightning App Builder, very cool!

Thanks Salesforce for filling my blog backlog for the next year or so! 😉

 

 


24 Comments

Image Recognition with the Salesforce Einstein API and an Amazon Echo

AI services are becoming more and more accessible to developers than ever before. Salesforce acquired Metamind last year and made some big announcements at Dreamforce 2016. Like many developers, i was keen to find out about its API. The answer at the time was “check back with us next year!”.

pipaWith Spring’17 that question has been answered. At least thus far as regards to image recognition, with the availability of Salesforce Einstein Predictive Vision Service (Pilot). The pilot is open to the public and is free to signup.

True AI consists of recognition, be that visual or spoken, performing actions and the final most critical peace, learning. This blog explores the spoken and visual recognition peace further, with the added help of Flow for performing practically any action you can envision!

You may recall a blog from last year relating to integrating Salesforce with Amazon Echo. To explore the new Einstein API, I decided to leverage that work further. In order to trigger recognition of my pictures from Alexa. Also the Salesforce Flow usage enabled easy extensibility via custom Apex Actions. Thus the Einstein Apex Action was born! After a small bit of code and some configuration i had a working voice activated image recognition demo up and running.

The following diagram breaks down what just happened in the video above. Followed by a deeper walk through of the Predictive Vision Service and how to call it.

amazonechoandeinstein

  1. Using Salesforce1 Mobile app I uploaded an image using the Files feature.
  2. Salesforce stores this in the ContentVersion object for later querying (step 6).
  3. Using the Alexa skill, called Einstein, i was able to “Ask Einstein about my photo”
  4. This  NodeJS skill runs on Amazon and simply routes requests to Salesforce Flow
  5. Spoken terms are passed through to a named Flow via the Flow API.
  6. The Flow is simple in this case, it queries the ContentVersion for the latest upload.
  7. The Flow then calls the Einstein Apex Action which in turn calls the Einstein REST API via Apex (more on this later). Finally a Flow assignment takes the resulting prediction of what the images is actually of, and uses it to build a spoken response.
    einstenandflow

Standard Example: The above example is exposing the Einstein API in an Apex Action, this is purely to integrate with the Amazon Echo use case. The pilot documentation walks you through an standalone Apex and Visualforce example to get you started.

How does theEinstein Predictive Vision Service API work?

revaflintsilverThe service introduces a few new terms to get your head round. Firstly a dataset is a named container for the types of images (labels) you want to recognise. The demo above uses a predefined dataset and model. A model is the output from the process of taking examples of each of your data sets labels and processing them (training). Initiating this process is pretty easy, you just make a REST API call with your dataset ID. All the recognition magic is behind the scenes, you just poll for when its done. All you have to do is test the model with other images. The service returns ranked predictions (using the datasets labels) on what it thinks your picture is of. When i ran the pictures above of my family dogs, for the first time i was pretty impressed that it detected the breeds.

EinsteinPredictiveVisionAPI.png

While quite fiddly at times, it is also well worth the walking through how to setup your own image datasets and training to get a hands on example of the above.

How do i call the Einstein API from Apex?

Salesforce saved me the trouble of wrapping the REST API in Apex and have started an Apex wrapper here in this GitHub repo. When you signup you get private key file you have to upload into Salesforce to authenticate the calls. Currently the private key file the pilot gives you seems to be scoped by your org users associated email address.

public with sharing class EinsteinAction {

    public class Prediction {
        @InvocableVariable
        public String label;
        @InvocableVariable
        public Double probability;
    }

    @InvocableMethod(label='Classify the given files' description='Calls the Einsten API to classify the given ContentVersion files.')
    public static List<EinsteinAction.Prediction> classifyFiles(List<ID> contentVersionIds) {
        String access_token = new VisionController().getAccessToken();
        ContentVersion content = [SELECT Title,VersionData FROM ContentVersion where Id in :contentVersionIds LIMIT 1];
        List<EinsteinAction.Prediction> predictions = new List<EinsteinAction.Prediction>();
        for(Vision.Prediction vp : Vision.predictBlob(content.VersionData, access_token, 'GeneralImageClassifier')) {
            EinsteinAction.Prediction p = new EinsteinAction.Prediction();
            p.label = vp.label;
            p.probability = vp.probability;
            predictions.add(p);
            break; // Just take the most probable
        }
        return predictions;
    }
}

NOTE: The above method is only handling the first file passed in the parameter list, the minimum needed for this demo. To bulkify you can remove the limit in the SOQL and ideally put the file ID back in the response. It might also be useful to expose the other predictions and not just the first one.

The VisionController and Vision Apex classes from the GitHub repo are used in the above code. It looks like the repo is still very much WIP so i would expect the API to change a bit. They also assume that you have followed the standalone example tutorial here.

Summary

This initial API has made it pretty easy to access a key part of AI with what is essentially only a handful of simple REST API calls. I’m looking forward to seeing where this goes and where Salesforce goes next with future AI services.