Blog Post

Apps on Azure Blog
5 MIN READ

Integrating AI: Best Practices and Resources to Get Started with Azure Cognitive Services

Yonet's avatar
Yonet
Icon for Microsoft rankMicrosoft
Apr 15, 2021

 

We use AI (Artificial Intelligence) integrated applications dailyfrom search engines optimized to find the most relevant content, to recommendation engines for streaming or shopping. During AI’s early years rising to popularity, improving applications with AI was only possible for companies with big budgets dedicated to research and expertspreventing companies that cannot effort an AI team to compete. Today AI is readily available for any product, without having to invest in research and development. There are open-source libraries that can help you train Machine Learning models like TensorFlow. With a fraction of the effort and the cost, pre-trained AI services are available to easily integrate into your applications, with APIs and UI based tools to train custom models for your specific use case. In Integrating AI series, I aim to help you decide if and how to integrate AI into your applications, get you started with Azure’s ready to use AI solutions, Cognitive Services and answer your most frequent questions when getting started.

  

Let’s start with these fundamental questions: 

 

  • What are the problems you can solve with AI?
  • What do you need to know before starting to build your solution?
  • How do you measure the success of your new AI features? 

 

What are the problems you can solve with AI? 

 

 

 

AI is a groundbreaking technology but not a magical solution for everything. It is important to know if you are adding value or solving an actual user problem. There are complex products like Wikipedia and Reddit that have a lot of information but use crowdsourcing and simple search to cater to unique needs without the help of AI. To make an informed decision, you need to start with your users needs. What are the problems they face? Is there a process that you can automize like filling expense forms that can be automated with Form Recognizer service? Send voice messages to your customers with updates using Speech Services? Do they make complex choices while using your product thacould be customized to your users with the use of PersonalizerDo you need to improve the usability of your application with voice interactions and Language UnderstandingIt is important to solve a real need for your users instead of assuming the solution that will be useful. User research is the best way to figure out the issues and a lot can be surfaced by user analytics. You can use Metrics Advisor AI service to detect anomalies and figure out future AI solutions as well 

Once you have a clear definition of the problem and define how to measure success, it is time to explore practical solutions You can read about the Azure customer stories and learn from their methods and design process. For example, read about BBC's customer story before you read about the technical story  of using Azure's Speech, Azure Bot Service and Language Understanding Services together to solve the customer needs they identified.

 

 

Most AI solutions can fall into two categories. The first major use case for AI is automating the mindless repetitive jobs. If thusers of an expense report or a hiring application need to type in information from a form or a receipt to your system, it is easily automated by OCR (Optical Character Recognition). Similar automations are possible for close captioning, translation, classifying images and automizing alert messages.  

 

The second category of AI solutions can be categorized as complex human decisions based on data. You could give your friends recommendations on what to watch next easily, knowing what they like, what they don’t like. For example, a streaming service with thousands of movies to choose from, cannot surface relevant content with simple filtering of the genres or release dates. It would take forever to choose what to watch by browsing unless you know the exact name of the movie. For a decision like recommendation among thousands or millions of results, AI might be better at recommending to your best friend, maybe even better than you over time. Understanding the language and intent of people is another example. A human can understand and classify a review as positive or negative easily. For machines to detect the same emotions, you must do more than detect certain words to get sentiment. 

 

What do you need to know before starting to build your solution? 

 
Some problems are easier to solve than others with AI. Experimenting with different tools to confirming your solutions is important. All the Cognitive services are easy to try out and here is how to do that: 

 

 

 

  • Once you create an Azure resource, you can see the code samples, API call examples and try out the Rest API end points directly on the Cognitive Services API Reference pages.
  •  
  • API descriptions

     

    Code Samples

     

    Request

     

 

Will your users love your solution? 

 

Scaling an application and polishing the user experience takes most of the development time. It is better to try out features fast and adjust before making the investment in perfecting the wrong experienceYou might assume an application flow that users are going to interact, but users can surprise you in their own creative ways of using your toolsPrototype your applications quickly and get user feedback early on. 

 

Power platform is one of the tools that allows you to create mobile apps that integrates important AI capabilities without writing any code. With power platform, you can easily deploy and share your prototypes, without leaving the platformUIAfter the free trial period, both training and using your AI models will cost but not as much as the development time of an actual app with AI and having to make major changes after the release. Check out some of the capabilities and use cases of AI Builder on Power Platform and how to train a custom vision model and creating a mobile app on Power Platform in this article.

 

 

There are other fast and easy options to add AI to your applications, without a big development investment, especially if you are adding the capabilities to an existing application. You can use a logic app to design an application on Azure platform to find twitter mentions of your brand and analyze the sentiment of the tweets. You can visualize the data on Power BI or your choice of visualization platform or tools.  

 

Once you integrate your AI solution, you can make the new AI features to a limited group of users and compare the effectiveness of your solution with your non-AI features. 

 

 

AI Show Live every Friday on LearnTV

 

Thanks!

 

Updated Apr 14, 2021
Version 1.0
No CommentsBe the first to comment