Artificial intelligence, or AI, is growing more and more in popularity, so it’s worth learning how to use AI services. AWS has its own set of 20+ services, ranging from Amazon Rekognition to Textract to Forecast. These are very useful because it allows developers to automate simple tasks without extensive machine learning knowledge. For example, with Amazon Lex, you can create chatbots that can help customers without human intervention.

The following diagram shows a list of every AWS AI and ML service.

At the bottom, we have our machine learning frameworks; these take the most skill and most time to develop, although they leave you with the most control and customization over your infrastructure.

The top of the chart shows the AI services mentioned previously, such as Amazon CodeGuru and Amazon Kendra. These require far less technical expertise and are much quicker to use than the ML frameworks, although you get considerably less control.

Finally, we have the machine learning services, which are a bit of a middle ground. You can use services such as Amazon Sagemaker if you have some machine learning knowledge and only need some control.

Every level of the AI/ML stack has its own pros and cons; every business will have different needs and will require different resources.

I also made a video briefly explaining every AI service and its purpose, so make sure to watch it if you’re interested!