If you weren’t at the AWS Alexa meetup hosted recently by OFX, not to worry, we’ve got you covered.
Aside from the great beers and pizza (probably the best meetup pizza I’ve ever eaten – thanks Made In Italy), the attendees had some really interesting projects on the go.
Well over 20 per cent of all mobile searches are done via voice. This figure will soar to well over 40 per cent in the coming years, as with the advancements of AI. It makes sense using voice as a consumer, particularly if you think about how many people still use their mobile phone or satellite navigation keypads whilst driving.
Businesses have already incorporated voice into many of their services. Think about your bank, “in a few words, please tell me what you are calling about”, you might say, “calling about my home loan.” Because you said “my”, they know it’s existing, and “home loan”, so it’s about a home loan. You are then seamlessly directed to the existing home loans department.
We chatted to the speakers from Amazon (Simon Elisha – Head of Solution Architecture ANZ), Octo (Eric Favre – Consultant), DevDiner (Patrick Catanzariti – Founder), and the hosts CloudTen (Malcolm Duncanson – Co-Founder) and OFX (Mike Watkins – Tech Lead) to gain insights into all things what Alexa.
What is Alexa? Are you using currently and if so what for?
Simon: Alexa is the voice service that sits within the Amazon Echo product. Alexa uses your voice, natural language and understanding of the context of what you are looking for and translates it into an Alexa Skill. Alexa Skills can do things like help you order a pizza, order an Uber, turn your lights off in your home, and activate your sprinkler system. AWS plays a key role in enabling Alexa and that’s exciting from a developer perspective. Across Alexa, teams use various AWS products, such as AWS Lambda for third party skills, Amazon DynamoDB for data storage, Amazon RedShift for data warehousing, and Amazon Elasticsearch for natural language entity resolution. Echo’s brains are all in AWS so it will continue to develop, get smarter, and add more functionality over time.
Eric: Alexa is a voice recognition and processing system provided by Amazon. I am currently using it mostly to explore its capabilities and try some use cases as experiments. The latest one was to trigger the deployment of some applications to production by voice control, with a confirmation code displayed on an LCD screen to have a deployment validation.
Patrick: Alexa is the first voice-enabled computing system to take hold with the public in a way that hasn’t been seen before. It’s a fascinating medium! My Amazon Echo is used for the most basic of things, like “What’s the time?” or “What’s news?”, but it also has begun to be a connected home controller for me as I turn on my bedroom light with an “Alexa, turn on my light” command too. The more I’ve used it, the more you really miss being in hotels and other places without having that voice assistant! It’s a weird feeling, but it’s the feeling which I think comes from the tech that is going to make its way into the mainstream.
Malcolm: Alexa is Amazon’s voice driven personal assistant that is making a huge impact in the US and the UK. We see voice driven applications as the next major shift in technology.
Aibiliti and Cloudten are developing Alexa skills for customers to augment their existing digital presence.
Mike: Alexa is one of a suite of maturing intelligent voice platforms from Amazon, comparable to Microsoft Cortana and Apple’s Siri. Implemented in the Echo, Dot Echo and Echo Tap (all by Amazon), these standalone devices sit in your home/office and allow you to control your home automation, ask for weather/news updates, order your groceries to be delivered etc, all by voice. What distinguishes Alexa from other offerings is the easy access and low cost to integrate it with any system you want. Here at OFX, we integrated Alexa into our core product, by building a custom application (called a “skill”) which queries our public APIs, allowing you to request a quote for FX transfer simply by asking “get quote, 500 AUD to GBP”. We are also considering placing an Alexa device in our front reception when unattended, so that visitors can say the person and nature of their visit, in order for the relevant staff to be notified via our internal staff directory and Slack.
Where do you see Alexa being most utilised over the next 12 months?
Simon: AWS is excited to see what the future holds – We are entering a golden age of AI and ML. We believe AI will revolutionise almost all aspects of technology – making it easier to do things that take considerable time and effort today like product fulfilment, logistics, personalisation, language understanding, and computer vision, to big forward-looking ideas like self-driving cars.[G4] The cloud has spurred a lot of researchers to innovate and experiment with new algorithms in deep learning and you will see more advances in reinforcement learning, auto tuning of models across a wide variety of domains in the future. As we move into the second half of 2017, and beyond, you can expect to see more news from AWS with regard to our investments, and innovation, in this area.
At the AWS Sydney Summit, we ran a Student Track mini hackathon where students were asked to design an element of automation for their lives – what is something they’d like to solve; what is something they’d like to automate. Some of their solutions were called ‘get me ready for my day’ which involved Alexa instructing their wardrobes to select their favourite outfits, turns on the shower to appropriate degree, check calendars for first class started etc. The students walked through AWS executives on a cardboard sketch the elements they wanted to design and presented it. They then had the chance to simulate this through the Alexa on the day. A key element of Alexa being a ‘thing’ right now is the fact that we [G9] are able to use machine learning, deep learning algorithms via the cloud to train the models required to make Alexa sound like a human interaction.
Eric: Mostly at home, either for domotic applications, such as turning on and off lights, changing TV channel, closing electric curtains… or for contexts that don’t require a display, such as listing a cooking recipe steps, playing music, taking phone calls…
Patrick: I think it’s still early days for it — it’ll largely continue to be used in people’s kitchens as kitchen timers, reading out recipes, playing people music and audio books… that sort of thing. However, I think these are the 12 months where adoption will grow and it’ll start to make sense for corporates and other groups to start taking notice of this trend towards a voice-enabled future.
Malcolm: We are seeing growth in enterprise circles where businesses are keen to offer their products and services via voice driven applications.
There are already some very useful skills published in the US, to allow you to order an Uber or your favourite Domino’s pizza, the volume of skills is set to grow exponentially as more businesses realise the value of engaging directly with the consumer via voice.
Financial services are potentially one area that could benefit hugely by allowing customers to check credit card balances for example.
Mike: Without a doubt, the home. Having used the Dot Echo at home for a few weeks it is incredible how seamlessly it integrates into your day-to-day; providing music, jokes, recipes and trivia all without having to touch a computer or smartphone.
How would you advise Engineers to stay up to date on Alexa?
Simon: Alexa is one form of a wider growing array of AI and machine learning platforms which AWS supplies. While Alexa is not available in Australia at this point the range of services and uses that are happening around the world is expanding exponentially.
Australian developers can still use the Alexa development kit and even do testing or experimentation without a physical device using: https://echosim.io/welcome?next=%2Fthat way, they can experience the cutting edge right now. Engineers can never stop learning about these services. There is an abundance of material out there to facilitate this that are either free or at a small cost. The best way to learns is hands on with the consoles, products and services AWS offers. Meet ups like this one are also incredibly helpful, as well as formal training courses.
Eric: First they should try it, understand what Alexa is (not) about, and they should keep an eye on Amazon announcements for upcoming updates, new capabilities. They should also look at the skills marketplace to see what can be done and foster new ideas, look at competitors’ products and compare, take part in hacking events, meetups, look at specialised blogs…
Patrick: Buy an Echo and use it! Look at the skills out there and try them out. Keep an eye out for tech news on the Echo, it’s constantly evolving and improving. The Amazon Echo team have a weekly email that goes out showcasing something new they’ve added — usually, it’s not that exciting and is America-centric, but sometimes it’s a pretty handy feature. I’ve got my own online course coming out around building a voice assistant that’ll feature turning it into an Alexa skill and I’ve got a weekly newsletter to help keep people up to date on emerging tech like this at DevDiner.com.
Malcolm: By signing up to be an Alexa developer on Amazon’s developer site, you’ll receive the latest news and tips from Amazon.
Mike: The Meetups (such as this one) and the AWS news feeds would be the best way to keep on the buzz and Pluralsite has an awesome course on programming custom skills on the Alexa platform.
What advice would you give to Engineers attending these meetups?
Simon: Engineers should use these meetups as a great chance to network and talk to peers on how they use AWS services, and collaborate on projects.
Eric: Get inspired, and don’t take anything for granted, try by yourself 🙂
Patrick: Just get out there and make things. Even small and simple things. Use existing services to make things easier to start with — Api.ai is a fantastic way to do this and saves a lot of time. It also helps give it some potential as a cross-platform voice-enabled app, rather than just an Alexa one.
Malcolm: Get involved with Alexa and AI, experiment, build and try out your ideas. And if you want to present at one of the meetups get in touch!
Mike: Challenge yourself and create your own custom skill, it’s easier than you think and makes conversations at Meetups more fun! You would be surprised how quickly you become motivated to share your skill with the community or even present at the next meetup.
And, a bonus question for Amazon – Does Amazon Echo use AWS IoT?
Simon: Alexa Smart Home Skills, which is built on AWS Lambda, is used by third-parties often alongside AWS IoT to implement their Smart Home Skills. Customers including Rach.io and Scout are using Alexa Skills and AWS IoT in this way.
If you would like to see an example of what can be done using Amazon Polly and Amazon Lex, check out this nifty little project on YouTube from a Cloud Architect at Amazon!
Check out all the photos from the night: