In the previous article of Amazon and AI, we have discussed some names of AI/ML tools by AWS. Now we will discuss each of them in detail.

Amazon Lex:

Lex is a service for making machines understand human speech and respond to it in a human-friendly manner. For example, Alexa, Chatbots, etc. Like you say anything to Alexa, like playing up a song from the playlist, adding up the playlist, creating a shopping list, etc. Alexa understands those languages you chose up for Alexa in its initialization phase, and whatever the users say, it understands and responds in a friendly manner.

Benefits:

• Lex also provides you with advanced deep learning functionalities like Automatic speak recognition (ASR ) to text. So at the backend, it's just recognizing human speech, words, statements and then converting them into the text and further performing the tasks for the user.

• It also uses NLU(Natural Language Understanding) to analyze texts and sense or guess the motives behind those texts intelligently.

• Lex also provides best-in-class user conversational experiences as the response comes directly from chatbots. For example, you have to book an appointment with the doctor, and you're using a website for it. The website has a chatbot or a phone call or message service which uses the chatbot. Now the user calls customer service to reschedule an appointment. Now the user is using the phone line, which further connects with the amazon lex service, used by the website underneath in the chatbot. Then amazon lex invokes the lambda function and looks up the customer's existing information in the database. Then there will be software for actually scheduling the activity.
• Further, the new scheduling data will be saved into the database once the new appointment date is updated. It also sends the confirmation message to the customer. This is how the complete flow takes place.

• Another use of amazon's lex is the provision of information bot services. For example, the chatbot on a particular website asks you if there is any specific date or time you want your appointment on and which doctor you wish to consult, which means it asks from the user and then identifies what the user wants. These practices have actually eliminated the need for human activity, which is actually the implementation of IA.

Amazon Polly:

Amazon Polly uses the advanced deep learning technologies to synthesize the speech that sounds like the human voice, like you are reading a story and want someone to narrate that whole story for you. This service is provided by Amazon Polly, it will actually start telling the entire story with a soothing voice, and you also have a choice of various languages. Thus amazon Polly allows you to make speech-enabled applications that work in different regions around the globe.

Benefits:

• As far as the benefits of amazon Polly are concerned:
• It provides you with natural voices, and you can choose from a number of accents with the excellent pronunciation

• The next benefit of amazon Polly is that it allows you to replay the voice after storing it. You are not charged with any fee for the repetition of that voice. You can play it for unlimited time and create the speech files as mp3 and save them on the clouds or use them offline as well.

• Polly allows real-time streaming as well. It provides you with fast responses. Like when you send a text to Amazon Polly's API. It returns the audio to you as a voice, the real-time streaming.

• You can also use Amazon Polly services to customize and control the speech outputs with Amazon Polly. You can also control the aspects of speech, e.g., pronunciation, volume, speed, etc.

• It charges you at a meager cost and is pocket-friendly as well. You just have to pay only for whatever you are using. So it is a low cost per character converter.

Amazon Rekognizer:

Amazon Rekognizer recognizes the videos, images, etc., and then helps you analyze them. You just have to give any sample video to Rekognition, and it will identify the image, texts, and people in any inapt content.

**_

Benefits:

_**

• The first and foremost benefit of amazon Rekognition is that it allows simple integration.
• It makes it easy to add visual analysis features to your applications with easy-to-use APIs that don't require any ML expertise.

• Then this service is based on algorithms that allow continuous learning, which means that it continuously updates itself by processing the new data sets and eliminating the previous patterns if required. It can recognize new objects and improve the ability to identify all the scenarios.

• It is a fully managed service as it provides consistent response times regardless of the volumes of the request you make. Then your application latency always remains constant even if the volume raises to even tens to millions.

• It is low cost as other AWS services as you only have to pay for the number of images and scenarios you want to process.