Ever wonder what a five day stay with more than 32,000 Amazon Web Services fanatics is like? Well, it is AWESOME!! AWS re:Invent 2016 was held in Las Vegas, Nevada from Nov 28-Dec 2. AWS re:Invent is the world’s premier cloud computing conference. This year’s AWS re:Invent focused on the Internet of Things (IoT), serverless computing, security, and Alexa. This conference was big. So, big that an 18-wheeler drove on stage during a keynote and stayed inside the exhibit hall for the rest of the event. For a sense of the scale of AWS re:Invent, here are some interesting stats: 697 sessions, 53 miles walked, more than 500 vendors in the Exhibit Hall, 1.8 million square feet of conference space, and more than 30 new AWS services and features announced.
What does AWS mean from an Audible perspective? Today, many Audible backend services currently utilize several AWS services such as SQS, SNS, DynamoDB, RDS, IAM and S3. At re:Invent, we learned about the latest new features associated with AWS services we currently use. More importantly, AWS re:Invent showcased new AWS services that several Audible Core Engineering teams are adopting for strategic initiatives. re:Invent provided an opportunity for Audible engineers to be exposed to the latest ideas in cloud commuting to help educate us to build the next generation architecture to power Audible services.
Ever wonder what a five day stay with over 32,000 Amazon Web Services fanatics is like? Well, it is AWESOME!!
AWS Lambda is a key AWS compute service that automatically scales in response to different events. It is a serverless computing service that allows a user to run code without managing servers. This service can be combined with other AWS services to build powerful big data applications. For example, every time a JPEG is added to S3, a Lambda function can be triggered to resize the image, and then the resized image can be put on S3. Or instead of the Lambda function resizing an image, the Lambda function could invoke AWS Rekognition, an image detection and recognition service, which returns a list of labels that identify the submitted image. As Lambda services become more and more complex, there needs to be tools that help manage these distributed systems. For such tools, check out AWS Step Functions, which provides the ability to coordinate and visualize complex workflows. The possibilities of how to combine Lambda with other AWS services are endless. Think big.
Alexa skill development was another large topic at AWS re:Invent. AWS Lambda is the key to Alexa skill development. In terms of software development, there are many questions that are still being asked. What does it mean to test voice? How do you design a voice interface? How does security impact such applications? These are all questions that are becoming relevant to the voice-space and need to be considered when developing voice-based applications.
One awesome session that I attended was the “Alexa MagicDoor”” skill. This session was about the development of an Alexa skill and the issues and problems encountered along the way. One of the more fascinating aspects was that this skill uses Alexa to read stories and during the story, Alexa asks the user to decide how the story should continue. For example, should the princess walk into a scary castle or run away? The MagicDoor Alexa skill also integrates sounds into the story like a door creaking or a wolf howling via SSML audio tags. These sounds help to enhance the Alexa experience because the Alexa voice can be somewhat robotic. Audible already rocks in terms of its audio experience and content; adding additional Alexa experiences that allow users to direct a story could be an interesting twist!
Check out some of the awesome sessions I attended (links are YouTube videos of the session):
- Transformations themed keynote by Amazon CTO Werner Vogels
- Amazon CloudWatch Logs and AWS Lambda: A Match Made in Heaven
- Another Day in the Life of a Netflix Engineer
- IoT State of Union
- From VUI to QA: Building a Voice-Based Adventure Game for Alexa
- Hold the Infrastructure: Getting the Most from AWS Lambda
Overall AWS re:Invent was a great experience. I am thrilled that I attended and I look forward to applying what I learned to my work at Audible. I definitely have more insight into the scale of AWS, a deeper understanding of the dev-op tools available, more interest in how Alexa will infiltrate our world, and how to better architect applications at scale.