Amazon Web Services (AWS), the leader in public cloud infrastructure now has more than 200 fully featured services, including compute, storage, databases, networking, analytics, robotics, Internet of Things (IoT), mobile, security, hybrid, virtual and augmented reality (VR and AR), media, application development, deployment, management, and machine learning and artificial intelligence (AI). For the latter, the message is clear: AWS wants to democratise ML technologies.
AWS has the most comprehensive set of AI and Machine Learning services for all skill levels. The most well-known is arguably the platform Amazon SageMaker, a fully managed service that removes the heavy lifting, complexity, and guesswork from each step of the machine learning process, empowering everyday developers and scientists to successfully use machine learning. Since AWS launched SageMaker in 2017, the company has added more than 150 capabilities and features, and already in December 2020 at that year’s re:Invent – when the first machine learning keynote took place – the message was simple.
As SiliconAngle put it, the company’s ‘overall aim is to enable machine learning to be embedded into most applications before the decade is out by making it accessible to more than just experts.’
With the AI & Big Data Expo, taking place in Amsterdam on September 20-21, AI News spoke with Felipe Chies, senior business development manager for AI and ML for the Benelux at AWS. Chies has strong experience in the field, having co-founded semiconductor startup Axelera AI, which has since been incubated by Bitfury.
Chies is speaking on the subject of accelerating innovation with no-code and low-code machine learning, and AI News spoke with him about key use cases, industries, and the different AWS products:
AI News: Tell us about the overall AWS ML and AI product set, how you talk about them with clients and how they help democratise machine learning.
Felipe Chies: We are very proud to have the most robust and most complete set of machine learning capabilities, and at AWS, we always approach everything we do by focusing on our customers. We think of our machine learning offerings in three different layers. First comes Frameworks and Interfaces for machine learning practitioners. These are people comfortable building deep learning models, working with deep learning frameworks, building clusters, etc. They can get extremely deep. Secondly the middle layer makes it much easier and more accessible for developers and data scientists to build, train, tune, and deploy machine learning models today with Amazon SageMaker. And last, Application Services, which enable developers to plug-in pre-built AI functionality into their apps without having to worry about the machine learning models that power these services. Many of our API services require no machine learning for customers, and in some cases, end-users may not even realize machine learning is being used to power experiences with services like Amazon Kendra, Amazon CodeGuru, Contact Lens for Amazon Connect, and Amazon HealthLake. The services make it really easy to incorporate AI into applications without having to build and train ML algorithms.
How does that help to democratise?
If we want machine learning to be as expansive as we really want it to be, we need to make it much more accessible to people who aren’t machine learning practitioners. Today, there are very few of these experts out there. So, when we built Amazon SageMaker, we designed it as a fully managed service that removes the heavy lifting, complexity, and guesswork from each step of the machine learning process, empowering everyday developers and scientists to successfully use machine learning. SageMaker is a step-level change for everyday developers and data scientists being able to access and build machine learning models.
To further democratize machine learning, we launched Amazon SageMaker Canvas, which enables business users and analysts to generate highly accurate machine-learning predictions using a visual point-and-click interface—with no coding required.
AI: How sophisticated does a customer of AWS have to be to use your AI/ML tools?
FC: AWS wants to take technology that until a few years ago was only within reach of a small number of well-funded organizations and make it as broadly distributed as possible. We’ve done that with storage, computing, analytics, databases and data warehousing, and we’ve taken the exact same approach with machine learning. We want it to be as broadly distributed as possible.
AI: What are the common use cases and industries that you see, and how can you help?
FC: Today, more than 100,000 customers use AWS Machine learning. One example of an industry where we see a lot of usage is manufacturing; and supply chain. With what has happened in the world most recently, there are many challenges in the supply chain area – so being able to forecast demand is very important. Customers ask us; ‘how can you help us to anticipate changes, to anticipate demand, to save cost to make our customers happy and deliver on time?’ Those kinds of things are common. For manufacturing, predictive maintenance, quality control – those are easy use cases to apply machine learning. For predictive maintenance, you can use computer vision to do quality control and more inspection. In marketing and sales, it is again forecasts. Forecasts are an area where it is easier to understand the value it brings to the business.
AI: What are the key roadblocks to ML adoption in your opinion and why?
FC: Many of the organisations I talk to already have a machine learning mindset so that is not a problem. One of the biggest challenges nowadays is the backlog of human resources– there’s just a lot to do for the development teams. One way to solve it is to get more people, but that’s another challenge – there’s just not enough specialists – it can be data science, machine learning, engineering – it’s really hard to find the people in the market.
This is really where the democratisation of machine learning comes in. Why not enable more people in the company to do machine learning? Instead of having only data scientists and machine learning engineers, why not also business analysts, or finance, or marketing people? An example of this is a tool like Amazon SageMaker Canvas. It enables business users and analysts to generate highly accurate machine-learning predictions using a visual point-and-click interface—with no coding required.
AI: What would you like attendees at the AI & Big Data Expo to learn from your keynote presentation?
FC: There are people who think maybe machine learning is something out of their reach, they need to go and send a requirement to the data science team and wait for weeks. This is not really the case – they can get started in a few minutes. This awareness that people can use machine learning nowadays without needing to know about it, how to build models – that is a key take away.
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London.