Well , it will be always better to have a friend or assistant who is always with you to help you. if the assistant is a computer , then it is better right. If my computer responds to me and talk to me feels better. Now a days our computers becoming more smarter. As part of smarter technologies , today we are going to look about Amazon Alexa or Amazon Lex.
Amazon Alexa is a single platform or complete package service available to build our own conversational bots, mobile and web applications. Lets call our bot friend as Lex. Lex uses the deep learning technologies like the ASR-Automatic Speech Recognition and NLU-Natural Language Understanding. Lex is a complete single managed service , where the tie required for the configuration was reduced. Since Lex has inbuilt NLU processing , we can use simple words in our Lex console while designing itself. Once the design was done and published. Lex will take care about our input and respond to the end customers.
Amazon lex has the capability to Connect ,FB messenger,Slack and Twilio. if we talk the same from the ground of Amazon, Amazon Lex has the capability to connect with Aws Lambda,AWS Mobile Hub,Aws CloudWatch. We can use Amazon DynamoDB and Amazon Cognito services for the proper functioning of Lex.
Let’s look how amazon makes different concepts together on creating a conversational bot.
Amazon considering Bot as a component having the capability to understand the end users need and responding back to the end user at the earliest. Amazon completely believes on fast transaction makes end user happy and satisfied. This can be achieved by setting a Context/Intent to chat, for example, i need to set a conversational platform with my bot to explain about my required like book for a movie ticket, raise an issue. The word intent will have different meaning in Natural Language Understanding(NLU) process.
For setting up a context , we can’t expect the end user to give intent alone as a keyword. For example: book bus ticket, here this is an intent, but real time the inputs will not be like intent alone. it will be like “i want to book a bus ticket” or “i want to travel from X to Y on 01 June 2017 by bus”. Here these are collection of words referring a single intent. Amazon is referring this as Utterances. The Amazon Lex development platform uses NLU Techniques to understand better about end users requirements.we can design the Lex , by promt / asking new questions to understand more on end user requirement / intent / context. Once the Intent was achieved, Lex will the use Lambda functions to support or fulfill our business logic based on user intent.
OMG , So many concepts, building my bot will be complex and difficult. Never building your bot in Amazon Lex is very simple. you can build your bot within 10 minutes.For this , you need a valid amazon account. Initially for 10,000 text requests and 5,000 speech requests available for each month at a free of cost for every month , Later we need to pay $4 for each 1,000 speech requests and $0.75 for every 1,000 text requests. Amazon Lex service will be available in preview format under the US East (Northern Virginia) Region. Let’s build our life as simple as possible with Bots, Happy Bot development.
To start with Amazon Lex development , refer the following link – LEX