AWS Specialty Alexa Skill Builder

I recently wrote the AWS Certified Alexa Skill Builder exam. I would be sharing my preparation journey to help others interested in taking this certification.

AWS recommends at least 6 months hands-on experience building Alexa skills using the Alexa Skills Kit. This is the most focused AWS exam, it is only concerned with building Alexa skills. It is also the easiest of all the specialty exams. You can take this as your first AWS exam if you have experience building skills.

The exam consists of 65 questions from 6 domains.

Domain 1: Voice-First Design Practices and Capabilities  

1.1 Describe how users interact with skills 

1.2 Map features and capabilities to use cases 

You should always remember, Alexa takes a voice first approach. Display is secondary.

Going into the exam, you should know the following:

·         Different ways users can invoke skills

·         Requirements for an invocation name

·         The characteristics of pre-built and custom skills and where they are used

·         Different utterances users can make when using a skill. Remember, it is best to define as many utterances as possible for intents

·         Design considerations for devices with display. There are limitations when using the Echo Spot

·         Events and directives

Domain 2: Skill Design

2.1 Design and develop an interaction model 

2.2 Design a multi-turn conversation  

2.3 Use built-in intents and slots 

2.4 Handle unexpected conversational requests or responses 

2.5 Design multi-modal skills using one or more service interfaces (for example, audio, video, and gadgets)

This section has the most questions and focuses on the “frontend” part of skills. Features tested here can be found in the Build section of the Alexa developer console. You should know the following:

·         The relationship between intents, utterances and slots

·         Slot types. Pay special attention to Search Query

·         Built-in intents and how to extend them

·         Design Patterns for situational design– Personalization, Availability, Relatability and Adaptability

·         Steps in designing an interaction - Start with scripting the conversations

·         Implementing skills for different locales

·         Multi-turn conversations – Slot Elicitation, Slot Confirmation, Slot Validation and Intent Confirmation

·         Using the Dialog interface for auto-delegation. You should also understand how to handle dialogs manually

·         The purpose of the Fallback Intent and Help Intent

·         Features of display templates and cards

·         Required and optional intents for different types of skills

Domain 3: Skill Architecture

3.1 Identify AWS services for extending Alexa skill functionality (Amazon CloudFront, Amazon S3, Amazon CloudWatch, and Amazon DynamoDB)

3.2 Use AWS Lambda to build Alexa skills

3.3 Follow AWS and Alexa security and privacy best practices

Know the use cases for the following services when building skills:

·         Lambda

·         S3

·         Cloudfront

·         Cloudwatch

·         DynamoDB

Domain 4: Skill Development

4.1 Implement in-skill purchasing and Amazon Pay for Alexa Skills 

4.2 Use Speech Synthesis Markup Language (SSML) for expression and MP3 audio 

4.3 Implement state management 

4.4 Implement Alexa service interfaces (audio player, video player, and screens) 

4.5 Parse Alexa JSON requests and provide responses

What you should know:

·         SSML structure and what each tag is used for

·         How to implement In-Skill Purchasing (ISP) and Amazon Pay and their use cases.

·         The requirements for ISP when using child-directed skills.

·         The types of products supported by ISP

·         The levels of state management i.e. application, session and user levels

·         The various interfaces that can be implemented for skills eg. VideoApp Interface, PlaybackController Interface and Display Interface

·         Actions that can only be performed using the ASK CLI

·         JSON requests sent by AVS and responses received from skills

·         Permissions required by skill and APIs for collecting user information

·         Skill types eg: Music Skill,List Skills and Flash Briefing

·         Requirements for  self-hosted skills

Domain 5: Test, Validate, and Troubleshoot

5.1 Debug and troubleshoot using Amazon CloudWatch or other tools 

5.2 Use the Alexa developer testing tools 

5.3 Perform beta testing 

5.4 Troubleshoot errors in the interaction model

You should know the following for this domain:

·         Validate slot values

·         Testing tools including the Alexa simulator, Echo devices, utterance profiler, Tone & Voice feature in the developer console

·         How beta tests are performed

·         Common errors when developing skills

·         Debugging and Troubleshooting using JSON responses from your skill

·         Debugging and Troubleshooting using Cloudwatch

Domain 6: Publishing, Operations, and Lifecycle Management

6.1 Describe the skill publishing process 

6.2 Add and remove users in the developer console 

6.3 Perform analysis of skill analytics in the developer console

6.4 Differentiate among the statuses/versions of skills (for example, In Development, In Certification, and Live)

You will find questions relating to publishing skills and managing live skills. You should know the following:

·         The skill publishing process and checklists

·         Team roles supported in the developer console

·         Versioning your skills backend

·         How Alexa manages Live and development stages of skills

·         The metrics Alexa provides for skills

·         Policies skills must comply with

·         Elements of skill manifests

Conclusion

With hands-on experience this exam is by far the easiest of the AWS specialty exams. Take time and prepare well. Understand the JSON interactions between the various interfaces. All the best!

Resources Used

A Cloud Guru

Whizlabs (Practice Test)

ASK Documentation

Exam Readiness: AWS Certified Alexa Skill Builder – Specialty

Alexa for Developers

Designing for Conversations