Power Platform

Speaker Information

Amey Holden CX Specialist in the Dynamics and Power Platform

Workshop Schedule

This workshop was ran on Microsoft Teams and set out to cover:

• Microsoft's Power Platform

• AI Builder

• General knowledge on Microsoft

The Power Platform

The Power Platform has applications that are centered around Azure, Microsoft's Cloud Platform. The Power Platform provides four unique solutions to empower businesses to tackle different problems that each organisation faces. These are also known as the four A's of the Power Platform.

1. Assist with chatbots

Instead of needing to hire a machine learning expert or AI specialist, businesses can build, train and deploy chatbots without requiring any code. These chatbots can then be used to assist customers or employees with handling certain processes or answering common questions.

2. Automate the work

In all lines of work, there are tasks that can always be automated. Whether it be entering data into a spreadsheet or sending out identical emails to clients, all these flows can be automated using Power Automate. Again, Power Automate provides automation with a drag-drop interface that is easy to use and doesn't require any technical expertise to get started.

3. Analyse the data

Now that you have your data, you want some analysis done on it to extract the important information from it such as trends and anomalies. PowerBI takes care of data processing and analysis, providing the ability to generate charts, graphs and dashboards from your data.

4. Act on the data

With the data in your back-end, now you'll naturally want some way to interact with the data and perform some actions through an application. Customers can now use the Power Apps solution to create and deploy prototype web applications for custom actions.

AI Builder

Did you know? The 'Not Hotdog' model from Silicon Valley contains over 250 lines of code?

AI Builder allows you to infuse AI into your apps without writing any code. The types of AI you can take advantage of falls under two categories; pre-built and custom models

Pre-Built Model	Custom Model
Like a pre-trained dog	Newborn puppy
Has a specific job	Need to train it to learn how to do everything
Already knows what to do	Doesn't know anything
Performs a functionality which you can utilise	Can be refined and attuned for all your unique needs

Types of Custom Models

1. Classification

• Train model with data to categorise the data

  • E.g. mailing list to tell the AI Builder to tag and route the email to the right people

• It can then go to help suggest tags for you

• Uses:

  • Detect spam

  • Remove comments

  • Route requests to the right people

2. Entity Extraction

• Categorise news/blog articles for content discovery

• Running your data into there can help tag those things for you

• useful for recommendations and personilisation

3. Form Processing

• Take forms and extract into data elements

• Help with handwritten to process the information in the form and that can be processed and sent to a database

4. Object Detection

• Train to detect objects

  • Check number plates

  • Count people in crowds for security

5. Prediction

• Feed it historical data

• It will anaylse the data and help to predict the future outcomes based on trends in the data

• Predict business outcomes and plan for such accordingly

General Knowledge Quiz