Amazon Lex is the behind the scene technology of Amazon Alexa device natural language understanding and speech recognition capabilities. Amazon recently announced the public availability of Lex for building voice recognition and NLP capabilities to mobile devices, web apps, and chat services. Now building all kinds of conversational applications are easier, including #Chatbots for your existing web applications.
As shown in the diagram, users can interact either via voice or text and chatbot fulfills requests by triggering AWS Lambda functions. Lambda can perform the requests like booking, inquiry etc. and respond back to the user. It can also perform other actions like sending notifications, storing data, invoking other services via APIs and many other functionalities that are possible on AWS platform.
Let us have a look at 3 main reasons why Amazon Lex is the recommended choice for building chatbots:
- It is a fully managed service: Lex runs on Amazon infrastructure, thus making it easier to scale based on the demand, hence no operational overhead.
- It is powered by Amazon Alexa: Lex is powered by the same deep learning technologies that power Amazon Alexa and lets you build natural language chatbots.
- It is integrated with AWS Lambda, a service that lets you run code without provisioning or managing servers. Lambda enables you to write and run logic for your chatbot using serverless compute.
Here are two reference tutorials explaining how to build chatbots using Amazon Lex:
The code for Part 1 and Part 2 is located in Github repository.