Azure Cognitive Services

Azure Cognitive Services is a cloud-based pre-built AI and is a fully managed Microsoft platform; it simplifies and speeds up the creation and deployment of AI solutions.

Azure Cognitive Services provides pre-built ML models that allow developers to enable speech, hearing, understanding, and reasoning in their apps while also bringing in live data. This differs from Azure ML, which is based on requiring you to create models trained on your data.

These pre-built models can be accessed by developers via the Cognitive Services API; this functionality can be added with some lines of code. This is a developer skillset and in contrast to the Azure ML service, you do not require data science or specific ML or language/framework knowledge to get value from these services; think of this as an ML-as-a-Service approach.

Azure Cognitive Services is comprised of the following:

Vision: This service provides identification and recognition capabilities for analyzing pictures, video, and other visual content.
Speech: This service provides text-to-speech and speech-to-text conversion, language translation, and speaker recognition and verification.
Language: This service provides pre-built scripts for processing natural language, key phrase extraction, sentiment analysis, and user recognition.
Decision: This service provides personal recommendations, content moderation, and abnormality detection.

This section looked at Azure Cognitive Services. In the next section, we will look at Azure Bot Service.

Azure Bot Service

Azure Bot Service is a cloud-based, fully managed platform that allows you to create intelligent enterprise-grade bots. Essentially, a bot is a software program that is designed to perform a particular automated task. Bots use AI to learn human activity and behaviors and interact through speech and text. The most common example of this is a chatbot, which uses conversational AI to help engage with customer service scenarios.

Azure Bot Service is provided as a PaaS service so that it can be quickly adopted with no need to create any underlying compute or infrastructure. Being built on top of Azure App Service, it has built-in capabilities regarding portal UI-driven simplicity, scale, and automation.

Web App Bot templates can be used to get started; you select the one that best meets your needs. The SDK language can be either C# or Node.js.

All this happens/is orchestrated for you in what is known as the Bot Framework; this streamlines the portal experience and allows you to build, customize, edit, and download your source code, view analytics, and so on. You can even do in-portal testing by asking your bot a question directly within the creation environment to see what response you get.

This section looked at Azure Bot Service. In the next section, we will look at Internet of Things solutions.