Deploy a Peltarion AI model in Microsoft Power Apps

October 11/4 min read

    Good news for all the Power App enthusiasts out there! You can now deploy a model you’ve built in Peltarion in a Microsoft Power App. With this connector, anyone at your company -- code background or not -- can build AI-powered apps that make your processes more efficient!

    Let’s look further at a sample use case.

    Use case scenario

    How do you empower your customer success team with AI? Well, first your team needs to know what the customer is talking about, which can be very difficult depending on how much information you have about the customer when they reach out. Let’s see how we can build a way around this!

    The problem

    Many companies are currently looking at how they can build the best engines for customer success. But interactions with customers are a delicate business, and reducing the number of questions asked to the bare minimum has proved to be a key factor in maximizing customer satisfaction.

    What can be achieved?

    Using natural language processing techniques, we can create a model that predicts what product the customer is talking about based on their initial description in the customer service system. 

    Customer: On one of the dresses I bought from your website last month the lace overlay had a tear in it when it arrived. 

    Rep: I’m sorry to hear that! Would you like us to send you a new Lily summer dress to arrive later this week or would you prefer a refund? 

    Customer: I would like a new one please. 

    Rep: All right, no problem! I’ll put one through straight away. Just send through the broken one to our returns address so we can recycle it. 

    Customer. Thanks! That’s really helpful. 

    Rep: No problem! Have a nice day

    Our Power App connector lets you build AI-powered apps to make your processes more efficient.

    Opportunity for deep learning

    Using data from previous interactions with customers, where a product label has been added manually, we can train a model to make accurate predictions about what the customer is talking about from their first few messages with the customer support team, saving valuable time and making for more happy customers! 

    How is the AI model implemented?

    The most obvious implementation of a model like this is to integrate it with a customer service system, to automatically let your customer success reps know what product the customer is referring to without them having to ask them. But a model like this can also be used for analytics purposes. For example, if you want to find out if a particular product is leading to an unusually large number of customer support tickets, and then be able to quickly resolve the problem!

    Data requirements

    The model is trained on past customer service data with labels containing the product the customer was talking about. 

    Try building it yourself

    Try out this tutorial to build your own customer complaint tagging system using Peltarion’s AI platform and Microsoft's app-builder Power Apps!

      Let us help you explore a use case

      Johan Hartikainen
      Sales Lead
      Peiman Momeni

      Applied AI & AI in business topics

      02/ More on Applied AI & AI in business