Recently, I wrote a nifty little article about Workspace ONE Mobile Flows which was super fun and appealing. Mobile Flows is not a new concept by any means, but VMware delivers it in an exciting way. REST APIs and OAuth do a wonderful job of delivering notifications and/or approvals to the Intelligent Hub giving us a one-stop-shop and creating seamless harmony. It’s VERY NICE! Now, we turn our attention to the exciting Virtual Assistant that VMware unveiled over the last year powered by IBM Watson who is trying to make a comeback of their own by revamping pricing and carving out a special place in the hearts of enterprises everywhere.

Turning our attention to Virtual Assistant, we will cover configuring Virtual Assistant in Hub Services, Chatbots, configuring Chatbots in the console, setting up IBM Watson, and finishing things up with a nice demo of the immersive experience delivered with Virtual Assistant. I’ve found this to be the most exciting thing that I’ve gotten to work with Workspace ONE and I know you will see how fun it is once we get there.

Configuring Virtual Assistant in Hub Services

I think that configuring Virtual Assistant is the logical place to start. The first thing that you will need to do is sign up for an account at the IBM Cloud. Once you have your account, you can signup for Watson Lite and setup the Virtual Assistant. The initial setup is pretty easy which you will see from the video below. As mentioned in the video, go and grab the Virtual Assistant sample here.

Depending on how things went, you may need to click “Swap Skill” and select the Hub Assistant, but I’ve seen it go differently in different tests. The idea is to make sure you have the right stuff assigned.

Once you setup Watson, you can go and grab the settings that you need for setting up the Virtual Assistant in Hub Services. You go into “Settings” in the Assistant page (that means ignore the erroneous text you see in the Hub Services portal).

The information you need for Hub Services is below:

Simply, input the data from these 3 fields and save them in the Virtual Assistance area. The big key is to make sure the Assistant URL is only the base URL, which can easily trip you up:

Elementary My Dear Watson

Now, we discuss the different elements of Watson before going into setting up the Chatbots. We will focus on 3 different terms inside of Watson and some examples: (1) Intents, (2) Entities, and (3) Dialog. It’s good to understand how they work as you move forward and want to create and enhance what VMware gives you.

Watson Intents

Intents are basically an object that analyzes what someone is trying to do. The idea is you teach Watson user intents in the form of the questions/things that a person says to the Virtual Assistant. Intents are mapped to dialogues which we will discuss in a bit. The idea is that you create an intent e.g. #vmw_FILE_GENERAL_TICKET which is a built-in intent for creating tickets in your service desk software e.g. ServiceNow. From there, you build examples for that intent. This is how we teach Watson and help Watson evolve. Watson will evaluate what people say and map that to an intent, which can then be passed to an action a.k.a. “Dialog”

The brilliance with Watson is you can teach Watson how to handle questions, work with the jargon of your organization, and handle the multitude of things that your people throw at you. Let’s check out a short demo on Intents.

Watson Entities

I’ll only spend a minute or two on Entities as they’re not really used in the samples that are out there currently. Entities are interesting in that they help you work with users in interesting ways. Entities are used to identity parts of what people say like names and dates. The system itself has built-in entities that can be used in your intents to understand numbers, terms, etc. to work with them effectively:

A great example that I found on how you can work with entities can be seen below. This example shows how entities are able to interpret days properly when asked a question:

Watson Dialogues

A Watson Dialog is basically a listener. You tell it to listen for a specific intent (which is passed in the header to Watson) and respond with a particular response. You can see what this looks like below:

Chatbots

The Virtual Assistant will leverage connectors (specifically Chatbots) to handle certain requests. The Chatbots will listen for the intents and pass them to the Chatbot provided that Mobile Flows are enabled and your Chatbots are configured correctly. By definition: “Chatbots are a piece of AI software that can simulate a chat with a person and talk and interpret them like a human being.” You can see how Chatbots interact with the Intelligent Hub and Watson below:

If you look at the source code for the ServiceNow connector, you can see how it looks for WorkflowId’s that match the Intents that you saw earlier (which will be passed in the header of the request):

package com.vmware.ws1connectors.servicenow.constants;
import com.fasterxml.jackson.annotation.JsonValue;
public enum WorkflowId {
CREATE_TASK("vmw_FILE_GENERAL_TICKET"), VIEW_TASK("vmw_GET_TICKET_STATUS"), ORDER_A_DEVICE("vmw_ORDER_NEW_DEVICE"), ORDER_LAPTOP("vmw_ORDER_LAPTOP"), ORDER_DESKTOPS("vmw_ORDER_DESKTOP"), ORDER_MOBILES("vmw_ORDER_MOBILE"), ORDER_TABLETS("vmw_ORDER_TABLET"); private String workFlowId; WorkflowId(String workFlowId) { this.workFlowId = workFlowId; } @JsonValue public String getId() { return workFlowId; }
}

We will see more in the demo how the Chatbots work to provide an immersive experience. VMware offers you a few out of the box for Workspace ONE UEM, ServiceNow, and LinkedIN Learning for example to deliver on the promise. One of the things that I really love about the sample JSON is that it checks for mobile flows and passes you to an immersive menu where you can select one of these chatbot options as you can see below:

Configuring Chatbots in Workspace ONE UEM

We will be setting up two Chatbots for our demo: (1) Workspace ONE UEM and (2) ServiceNow. In the last article, I showed you how to setup the regular ServiceNow connector so we can use the same OAuth token, which makes that setup a breeze. We will start with WS1 UEM.

Setting up the WS1 UEM Chatbot

Adding this Chatbot is super easy. The main key is making sure that your Base URL matches your console (that means CN not AS or DS):

Once you do that, it’s just a next, next, next party and that Chatbot is easy as pie. Yes, it’s that easy!

Setting up the ServiceNow Chatbot

Before, we get started we need to make one small change on ServiceNow. You will go into the Identity Provider for WS1 Access and click “Set as Auto Redirect IdP” because all Chatbots requires a direct SSO experience. Hitting your base URL must redirect to Access or it does not work period. One side note: after making this change it will take up to 15m for your Chatbot to start working:

The Chatbot setup is the same as I highlighted in my previous article with one exception, the ticket tables you need to set for ServiceNow, which you can see below:

Now, that you have two Chatbots set and ready to go, we can start our demo and you can see how the power of the Virtual Assistant comes together!

A Look at the Virtual Assistant

Now that we have done some nice work, we can check out a demo of the experience in the VMware Workspace ONE Virtual Assistant powered by WS1 UEM and ServiceNow.

As you saw, we can agree that this is a demo that is very compelling and exciting for users. Many of the pain points we struggle with are finally eliminated. In theory, this eliminates an entire level of admins and makes life much easier and pain-free.

Closing Thoughts

Workspace ONE Virtual Assistant has finally come to fruition and provides a nice competitive advantage. Finally, WS1 has a feature that cannot be categorized as free with Intune. Without question, cost is a concern as you need a higher tier license than you current have. That means it’s time to start converting those Horizon and WS1 customers to a single license so they can get this for free with just a nominal cost of $1400 per year per 1000 users (the Watson pricing) which is nothing. Imagine a large scale deployment of 10,000 users getting a Virtual Assistant for 14k per year. That is a no-brainer. You can do so much with OOTB connectors that you can’t use the code argument as a reason to not do it. Let’s be progressive and take that onboarding experience to an entirely new level with Workspace ONE and its virtual assistant.