Skip to content

Latest commit

 

History

History
41 lines (31 loc) · 5.18 KB

OVERVIEW.md

File metadata and controls

41 lines (31 loc) · 5.18 KB
Raw

In this tutorial, we're going to build and run an Alexa app (called an "Alexa skill") Here's a high-level view of what happens when you use an Alexa skill:

skill diagram

Let's start at the top right:

  1. You (the blue person) talk to an Alexa device (such as an Echo, shown in black). We'll imagine the skill is called "Dave," so you say something like "Alexa, ask Dave to tell a joke"
  2. The Alexa device sends your command to some Alexa-handling logic in the Amazon cloud. It splits up your command to understand that you're talking to the "Dave" skill, and gave it the "Tell Joke" command (a command for a skill is called an "intent", so this would be known as the "Tell Joke intent")
  3. Amazon has a filing cabinet somewhere, or a virtual filing cabinet anyway, that has information about each skill, including where the skill runs. So Amazon looks up the information for the "Dave" skill.
    • Alexa skills can be run anywhere you like, but it's more straightforward to run them in AWS Lambda than anywhere else. Since that's also free for any normal amount of usage, we'll use it rather than come up with and configure some other hosting provider.
  4. Now knowing where the "Dave" skill lives, the Amazon cloud calls Dave, wherever it may be. (In our case, in Lambda in the AWS cloud.) It tells the Dave skill that you've given it the Tell Joke intent.
  5. Dave decides what to do. In this case, it might tell Alexa to say "Knock, knock". It replies to the Amazon cloud with that instruction.
  6. The Amazon cloud tells your Echo device to say "Knock, knock"
  7. The Echo device says "Knock, knock"

As for the purple part of the diagram, you'll learn more about these parts later, but just to introduce them:

  1. You define an Alexa skill in the Amazon Developer Console (this is a Web page). It is all configuration information there -- the name of the skill, the things you can ask it or tell it to do, etc.
  2. We'll run the code for the skill in AWS Lambda, which we can set up using the AWS Console (another Web page). Once the code is set up, you can just push your latest code to AWS Lambda and it will run there.
    • AWS Lambda is just a place to run small chunks of code called Lambda functions. You provide the code for the Lambda function, and AWS runs that code for you, whenever something calls your Lambda function. (So there has to be something to call it or else it will never run.)
  3. To make these things work together, you need to point the Alexa skill configuration in the Amazon Developer Console to the Lambda function running in AWS Lambda. That means that when someone tries to use your skill with Alexa, Alexa will call the Lambda function and AWS Lambda will run it for you to tell Alexa what to say or do.

What you need on your Mac to work with this Alexa skill:

All of these things will be installed as part of the tutorial!

development diagram

All of these things will be installed as part of the tutorial!

  1. For starters, we'll use source code control from GitHub. It will have the "master" copy of your code, which can be shared between people or across several machines. The set of code for a single project is called a "repository", so GitHub has your repository in the cloud.
  2. On your machine, you'll have a local Git repository, with your working copy of the code. You can send your code changes from there up to GitHub, and (on a multi-person project) pull other people's changes from GitHub down to your machine.
  3. Inside that repository on your machine, you'll have your working copy of the code that should run in AWS Lambda. It will use some code written by other people and made available for your use -- these are called "libraries". And you'll have a copy of the helper tool called "Gulp" that can do things like send code from your local machine to AWS Lambda to run there.
  4. Elsewhere on your machine, you'll have the Visual Studio Code app, which you use to actually edit the code and configuration files for this skill.
  5. You'll also have some tools needed to work with JavaScript files. We're using JavaScript for this because it's well-supported, runs fast in AWS Lambda, and is useful to know in a world of Web apps.
    • Node runs JavaScript tools and apps on your machine
    • NPM is part of Node -- the Node Package Manager -- that knows how to install libraries and tools for a JavaScript project
    • Unfortunately, there are a large number of Node versions out there, that can sometimes conflict. We use NVM, the Node Version Manager, to control exactly which version of Node a given project will use. This is especially important because AWS Lambda uses specific versions of Node, and we don't want to use an incompatible version by accident.
  6. Finally, you need the AWS Command-Line Interface (AWS CLI) tool to work with AWS. You can inspect and update your AWS configuration that way. Gulp uses the AWS CLI under the covers to send your code to run in AWS Lambda.

Next Steps

The step-by-step tutorial starts with Running your first Lambda, which includes setting up all the tools we talked about above, and running a simple Lambda like in the top diagram.