- Each team should register to Alexa Developer Console:
- Go to: https://developer.amazon.com/alexa
- Click on "Create a skill"
- For testing Alexa Skills, each team need any of the following:
- A BYOD Alexa Device. If they choose this option, they should register the Alexa device using the same amazon account they used when registering to developer.amazon.com.
- A web browser that supports Audio Inputs (ie. chrome)
- They can use the Alexa Developer Console Tester
- or use https://echosim.io. Just register using the same amazon account they used when registering to developer.amazon.com.
- Each team should have a unique skill name to prevent accidental Alexa Start Session trigger.
- Each team will be provided their own AWS Account to use.
- We will be using the Cloud9 as our IDE and SAM (Serverless Application Model) to provision and deploy our AWS resources.
- We will be using N. Virginia (us-east-1) region for this hackathon. Always make sure that you are in this region when interacting with the AWS console, CLI, or SDK.
- Each team should be able program using any of these languages: Node.js, Java, Python, C#, or Go. The sample skills and scaffolding code provided are written in Python.
-
The theme of this hackathon is "Building Productivity Alexa skills". This could be something like:
- Meeting of the Minutes skill
- Meeting room booking skill
- A skill that voice back sales stats
-
While planning what skill to develop, make sure that you can deliver it within 3 hours.
-
Each team is allowed to use any other AWS services such as DynamoDB. This will give you extra points.
-
You can use any other third party API's. PROTIP: see if there's any interesting API's to play with in https://rapidapi.com/
- Once logged in to AWS console. Go to Cloud9 dashboard and create a new Environment with the following configuration. Make sure that you're in N.Virginia (us-east-1) region:
- Name: Any
- Environment Type: Create a new instance for environment (EC2)
- Instance type: t2.micro (1 GiB RAM + 1 vCPU)
- Platform: Amazon Linux
- Cost-saving setting: 30 minutes
- Network VPC: will be provided
- Subnet: will be provided
- Click Create Environment and wait for the Cloud9 instance to boot up.
- In Cloud9 IDE Terminal window, clone the hackathon materials
$ git clone https://github.com/adelagon/alexa-hackathon-runbook.git- We will be discussing how to build your own Alexa Skills on the section.
-
In Cloud9 IDE, go to python samples folder and open the app.py file that contains the alexa skill written in python.
$ cd alexa-hackathon-runbook/samples/python/alexa-random-chuck-fact/src
-
The provided example is a custom Alexa skill that give random Chuck Norris facts. At this point, we will spend some time to run through the well-documented code. This should give everyone an idea on how to write your own skill.
-
Focus on the Cloud9 Terminal Window and follow the next steps. Install dependencies:
$ cd /home/ec2-user/environment/alexa-hackathon-runbook/samples/python/alexa-random-chuck-fact/src$ pip-3.6 install -r requirements.txt -t .
-
Create an S3 bucket for deployment:
$ aws s3 mb s3://ph-alexa-hackathon-<team name>
-
Create a lambda deployment package:
$ sam package --output-template-file packaged.yaml --s3-bucket ph-alexa-hackathon-<team name>
-
Deploy the Application:
sam deploy --template-file packaged.yaml --stack-name alexa-random-chuck-fact-py --capabilities CAPABILITY_IAM
-
Go to AWS Lambda Dashboard and look for the Function named alexa-random-chuck-fact-*.
-
Copy the ARN of the Lambda function that was produced. This can be found on the Top Right corner of the Function detai page.
-
Login to https://developer.amazon.com/alexa and create a new skill. Put any value to the skill name and select "Custom" on Choose a model option and "Provision your own" on Choose a method to host as seen below then click "Create Skill":
-
Choose "Start from Scratch" when choosing a template. Then click "Choose".
-
On the right side panel of the Alexa Developer Console. Click "1. Invocation Name" to specify the utterance that will trigger the Alexa Skill.
-
Enter the Skill Invocation Name. (ie. chuck norris). It could be any value. This will trigger the
LaunchRequestHandler. Click "Save Model"
-
Now let's Add our Custom Intents. On the Left Panel. Click the '+ Add' Button right beside Intents:
-
Type AChuckFact in Create custom intent. Then click "Create custom intent"
-
Now add Sample Utterance as 'a random fact'. This will trigger the skill to provide another fact which is implemented in the
RandomFactHandler.
-
Now let's add Slots. Click "+" beside the "Slot Types" on the left panel. Name the Custom slot type as categories. Then add two slot values: science and music.
-
Add another Custom Intent named 'AChuckFactWithSlot' and use the Slots we have created in Step 14. The utterance should be 'a {categories} fact', using that kind of utterance links the Slots to this Custom Intent. This custom intent triggers the
RandomFactWithCategoryHandler
-
Add another Custom Intent named 'Crash'.
-
Now add Sample Utterances like: 'crash'. This will trigger the skill to simulate a skill bug in the
CrashHandlerwhich will then trigger theAllExceptionHandler.
-
Now let's add the Utterances for the built-in AMAZON.HelpIntent. You can use 'help' and 'help me'. This will trigger the
HelpIntentHandler
-
Add the Utterances for the built-in AMAZON.StopIntent. You can use 'stop' and 'goodbye'. This will trigger the
StopIntentHandler
-
Copy the Skill ID. We'll use this on the next steps.
-
Go back to AWS Lambda dashboard and select the Lambda function named: alexa-random-chuck-fact-* that was created when you did Step 5. Click "Add Trigger" on the Configuration page.
-
Select "Alexa Skill Kit" from the drop down of available lambda triggers.
-
Provide the Alexa Skill ID that you have collected in Step 20. Then click Add.
-
Go back to the Alexa Developer Dashboard and click on the "Endpoint" option on the left panel. Provide the Lambda ARN that you have obtained in Step 6 in the Default Region. Then Click "Save Endpoints"
-
The Alexa Skill Checklist should look like this.
-
You may now test your Alexa skill. Try the following conversations:
- "Alexa ask Chuck Norris a random fact"
- "Alexa ask Chuck Norris a music fact"
- Try saying "Crash" or "a science fact" or "help me" or "goodbye" during your dialogue with Alexa
- Whenever you change anything on your code, use sam package to package it and sam deploy deploy to AWS:
sam package --output-template-file packaged.yaml --s3-bucket ph-alexa-hackathon-<team name>sam deploy --template-file packaged.yaml --stack-name <stack_name> --capabilities CAPABILITY_IAM
- You can install any third party library. In python you can do this by adding an entry on requirements.txt file and running
pip-3.6 install -r requirements.txt -t .within the src folder. Be sure to repackage and redeploy when you are doing this. - For faster testing, you can use can configure Lambda Test Events using the Amazon Alexa event templates.
- Use AWS Cloudwatch to see the lambda logs.
- Have multiple tabs open on your web browser for both AWS console and Alexa Developer console. Better yet, split tasks amongst each team member.
- Use a good invocation name. The one that is easy to pronounce.
- Alexa Skills Kit - https://developer.amazon.com/en-US/alexa/alexa-skills-kit/learn
- Bult-in Intents List - https://developer.amazon.com/docs/custom-skills/implement-the-built-in-intents.html
- ASK SDK for python docs - https://alexa-skills-kit-python-sdk.readthedocs.io/en/latest/
- ASK SDK for node.js docs - https://ask-sdk-for-nodejs.readthedocs.io/en/latest/
- ASK SDK for java docs - https://github.com/alexa/alexa-skills-kit-sdk-for-java
- AWS SAM - https://aws.amazon.com/serverless/sam/
- AWS Cloud9 docs - https://docs.aws.amazon.com/cloud9/latest/user-guide/welcome.html
- Alexa Samples - https://github.com/alexa