Following the collaborative partnership between SLASSCOM and Amazon Web Services (AWS), a workshop was held on the 27th of June. The AWS workshop was hosted by Hatch, together with AWS. It gave participants the chance to learn about AWS fundamentals, Artificial Intelligence & Machine Learning services, IoT and a host of other topics.
The AWS workshop essentially had participants deploying a simple web application that enables users to request unicorn rides from the Wild Rydes fleet. By means of an HTML UI, users could indicate the location where they would like to be picked up. Once done, the backend of the application would process the request and dispatch a nearby unicorn. The application will also provide facilities for users to register with the service and log in before requesting rides.
Learning the basics at the AWS Workshop
Before getting started with making the application, participants were required to set up an AWS account. Those with accounts simply proceeded to log in and go on to the next phase while those such as myself went about creating an account.
With the account set up, the participants took part in the next steps. These included configuring the AWS Cloud9 IDE, configuring the AWS Amplify Console and AWS CodeCommit services.
From there, participants had to choose from either using AWS CodeCommit or GitHub to host the site’s repository. Depending on what the participants chose, the relevant steps were listed to configure each repository.
Once the repository is configured, the next step is to use the AWS Amplify Console to deploy the website that was just committed. If any changes are made to the repository, the Amplify console will rebuild the side with the changes.
Harnessing the power of AI and Machine learning through AWS
While the participants were diligently coding their way through the AWS Workshop, Ben Romano – Solutions Architect at AWS gave a brief overview of AWS AI/ML services.
Ben spoke about the differences and similarities between AI, Machine Learning and Deep Learning. Machine learning comes in a number of types. These range from unsupervised learning, to supervised learning and reinforcement learning. Amazon, for example, has been making investments in machine learning for the last 20 years. The mission for AWS is to put machine learning in the hands of every developers and data scientist.
Ben then went on to share a number of examples of the AI/ML services that Amazon uses. For example, Amazon Rekognition is Amazon’s image recognition service. The service can be used for scene detection, face detection, and object recognition. An important note that Ben shared was that all this happens in real time. There’s no batch processing involved. You feed an image to Rekognition, and it will identify objects on the spot.
Taking things a step further is Amazon Rekognition Video. As the title suggests, this service allows you to detect faces and objects in video. Once again, this happens in real time. By using footage from a CCTV camera, ben demonstrated how Amazon Rekognition Video is able to differentiate between humans, animals, and objects and even predict the movement of objects and people.
With Amazon Transcribe, you have the ability to turn audio into text. The service uses machine learning and AI to carry out automatic speech recognition. This is particularly useful if you are looking at documenting audio files for later reference. Amazon Polly, on the other hand, converts text into lifelike speech. This allows you to develop applications that talk. It uses complex deep learning technologies to synthesize speech that sounds like a human voice.
If you’re looking for Conversational Bots, AWS has you covered. Amazon Lex uses voice and text to build conversational interfaces into any application. It uses the same deep learning methods of automatic speech recognition (ASR) to convert text and to understand natural language and intent of the user.
Learning all about the AWS IoT Core
The Internet of Things or IoT is all the rage these days. But as we learned at the AWS Workshop, the trick with IoT devices is that they are complex and multidimensional. The Internet of Things comprises of three pillars or areas. They are Devices, Cloud and Intelligence.
IoT Core is just one of the many AWS IoT services. The presentation covered exactly how AWS can assist you when developing an IoT application. Participants who asked questions were gifted with an IoT device. This device was used to carry out an IoT Workshop at the AWS Workshop itself.
The exercise was to simulate temperature fluctuations on the IoT device. An email would be sent if the temperature was below 25 Celsius. On the other hand, if the temperature was above 25 Celsius, an SMS would be sent instead.
Using AI/ML in the Real World
The last session for the day at the AWS Workshop was pretty self-explanatory. Ben Romano was back to talk about how Amazon uses AI and machine learning to streamline their operations. One such example is the personalized digital user experience.
This is something Amazon is very well known for. If you shop for something on Amazon, the next time you log in, you might see products that are related and recommended for you. This is the personalized digital user experience at play.
Ben shared that a user’s expectations are continually evolving. They want things to be about them. But doing this is difficult. You need to know who your audience is and how to reach them. In short, you need to engage them in the right place and time.
Ben then spoke about Amazon Personalize. This is Amazon’s personalized recommendation system. It is a machine learning service that makes it easy for developers to create individualized recommendations for customers using their applications. Ben went on to explain that it’s continuously improving and works on your data. It’s also extremely easy to use.
Ben’s next topic was about Amazon Forecast. This involves analyzing data that has been collected over time to come up with a forecast of about products. The key problem here is that accuracy is the most important factor in forecasting. If you are carrying out forecasting, you can use deep learning to increase forecast accuracy. But this takes a lot of resources. This is where Amazon Forecast comes into the picture.
With that, Ben’s session on AI/ML in the real world came to an end. Meanwhile, participants were actively engaged in the AWS workshop by either creating the website or playing around with the IoT device. At the end of the AWS workshop, not only had they learned about AWS and what it has to offer, but they also walked away with a number of cool rewards as well.