This post is of course about my first impressions using Copilot for Power Virtual Agents. It is also an excuse to quickly play around with it. I want to put this technology to a quick test as I did with Boost Conversation for Power Virtual Agent. I am very much an enthusiast of this technology and the productivity it can bring. So let’s put it to the test!
ChatGPT and OpenAI has taken the world by storm. Now almost everyone is talking about AI and especially generative AI. First DALL-E shows us we can generate images and art just by prompts then ChatGPT shows we can generate entire blogpost by doing the same. I find these AI models thoroughly impressive but what I find even more so is how fast Microsoft and other tech companies are incorporating these models into their products. Today I want to focus on one small piece of these announcement and give it my thoughts by asking myself the following question: When and how should you use Boost Conversation for Power Virtual Agents?
Let’s start off with a brief description of what Boost Conversation is for Power Virtual Agents. Boost Conversation allows you to wire up your PVA to a website of your choosing. Meaning it will read the content of that website and index it. And then tries to answer the questions a user ask to your Chatbot with data found on that website. For this it uses Azure OpenAI services paired up with Bing to access and index that website. So instead of using ChatGPT 3.5 to try to answer the question with the knowledge of the entire internet it will narrow it down. Is this a brilliant decision or a flaw by the product team?
In today’s data-driven world, it’s becoming increasingly important to separate data from logic when building applications. This separation not only helps keep your code organized and maintainable but also makes it easier to manage and work with your data. In this blog post, we’ll explore how to use Template Tasks to separate data from logic and make your applications more efficient and easier to manage.
I will do this by expanding my previous blogpost where we generate Tasks using a Custom Page. In that blogpost I show how to generate Tasks to help the user in their work. However for simplicity it did not contain a Description, or even all fields. Thus the Power FX logic looks pretty clean. If we want to keep it clean we can separate the data by using Template Tasks. By storing the data in Dataverse and using template tasks to generate tasks based on logic inside the Custom Page, you can create a more scalable and flexible workflow.
The first release wave of features for the upcoming year are available for us to look at! Always fun to digest what’s coming. And as always, it’s important to note that these plans are subject to change. Consider them as a general guide for what’s to come. Please remember that this is an evolution of the platform, not a revolution. Microsoft typically saves major announcements and game-changing features for their big events, so don’t expect any earth-shattering developments in this wave. Nevertheless, there are still some exciting updates upcoming that are worth taking a closer look at.
Now the first important announcement is about the planner itself. The Release Planner tool is now General Available and I encourage you to use it! Create your personal list of features you are exciting for and easily see what has changed since your last visit. Or you can create one to share with your team to keep track on new features you need in your solution. This is also an excellent example of dogfooding as it is build on top of Power Pages!
I can categorize my highlight list in 4 topics; Improvements to the UI, Maker productivity, Application Lifecycle Management and Power Pages. And if you read more, I will address them as such in this blogpost!
Sometimes you have a step in a process which is so complex you want to guide the user. You can guide the user by giving him steps of Tasks to do. But what Tasks do you present to your user? What if those steps of Tasks depends on variables that require human interpretation? We can create a Custom Page Dialog to Guide the user what Tasks to generate! That concept is exactly what I want to show you in this blogpost.
We will extend a process I have written extensively about. The Case of the Broken Screen. In one of the steps we sometimes have to coordinate repair work. And sometimes we have to coordinate sending replacement parts. All the while by personally informing the customer. To manage this we we will be using the Tasks set-up from this Blog post on our Case table. This allows us to relate Tasks to the case we are working on.
But what Tasks to work on? Let’s open the Generate Tasks Custom Page Dialog to create them!
When assigning critical Tasks to different users or teams, you probably want to know when they have finished them! You probably also do not want a mobile push notification disturbing you in your work. Luckily Model-Driven Apps provide a way to notify users! In this post I show you how to accomplish this using Power Automate and the In-App Notification center of Model-Driven Apps.
But what I find really cool, is that you can create a richer card for your notification. You do so by combining the body and data column of the notification table. In the body you can use some Markdown to style your message. In the data column we can add json to add actions in the notifications. Actions allow you to add navigation urls to for example a form in your app. Building these is a breeze with the In-App Notification Builder in the XrmToolBox.
Wait? Didn’t I already have a blogpost about vertical buttons for Power Virtual Agents? Yes, I do, but it is time to revisit! I will show you how to add Vertical Buttons to a Power Virtual Agents question. I will use the new advanced authoring canvas to do so. As you may know, previously we had to use the Bot Framework Composer for this and export the dialog back to Power Virtual Agents. In fact, I describe this solution in my most popular blogpost of the last year.
The new authoring canvas now is in preview. The authoring experience is rebuilt from the ground up. Now we can use a lot of great new features in the authoring canvas directly such as: posting Adaptive Cards, adding rich media like images and videos and even use Power FX throughout! I am excited about this direction. And in my humble opinion, it stays easy to use despite all these new functionalities! Keep in mind all functionality is still in preview and should not be part of your production Bots yet.
In my previous blogpost we created tasks in Dataverse which we can interact with using the Model-Driven App. However, sometimes the user we assign a task to does not log in to our application as often as we would like. So, we need to bring them to attention a task is waiting for them. An excellent way to do this is by automatically sending them a chat message and posting an Adaptive Card! In this post I will go through the steps to set this up.
This will not be a crash course on all the ins and outs of every detail of Adaptive Cards. But more of an inspiration style post to get you started with your Adaptive Cards. I will however, show you how to set up the Cloud Flow to post this card to Teams. Including the ability to process a response from that Card.
In this blogpost I will start with the basics of “Task based work”. A task system that consists of a framework to add and coordinate tasks from a Model-Driven Power App. It is not very complex to begin with, but I will add more complex functionality to it as we go. This functionality heavily depends on the integration scenario we want to implement. In this post I have summed up a few possibilities.
A quick reminder of what we are working on. The big theme of my blog currently is the “Case of the broken screen”. A lot of my blogpost are about this business problem. A loyal customer has dropped their new and expensive phone and his screen is cracked. It is up to us to resolve this incident to satisfactory of the customer. It’s time to send out a replacement screen!
The caveat though, is a different department handles shipping. Sending a handwritten note or an e-mail to that department might get the job done, but it will not create tangible data to work with. So now we want to set out tasks to the logistic department who are not using our application. We also want to be able to track those tasks, make sure they get done and keep our customer informed of the status. Eventually you would want automation in place to be notified at the right time.
You might have seen the new AAD User table appear in your environment. This table is available to us for a couple of months now. And you might wonder what it is and how you can use it. In this blog I will explain you what it is and how you can use this table to your benefit.
The AAD User table lists all users in your Azure Active Directory. It does not show AAD Groups. Did you know that the data of this table does not reside in your dataverse storage? It is actually a Virtual Table! Virtual Tables are tables that do not reside inside Dataverse, but instead reference a table from a different storage mechanism. It is incredibly powerful as it does not require you to copy data from another system. You can find if a table is virtual by looking at the type column of the table overview. So why is this interesting and what can you do with it?