AI, ML come of age at AWS re: Invent
February 2, 2018
Print this page
By Paul Barker
Las Vegas – The one guarantee delegates attending an AWS re: Invent can be sure of is a dizzying array of new product and service launches. In fact, during a single keynote speech from CEO Andy Jassy, at the sixth annual edition of the developer conference held late last year, there were no less than 21 different announcements ranging from new container capabilities to new IoT services the company said will bring machine learning to the edge.
“The reality is that when you’re making a big transformation like the cloud is, the longer you take to make it, the harder it is to execute because you get deeper into the hole of whatever you’re building and what you need to crawl out of,” said Jassy.
“This is not about skating to where the puck is. The puck’s been dropped and it’s right in front of you and you have to decide are you going to play or are you going to skate away and not play? For companies, there is a huge penalty for not playing because you have less capable technology than all of your competitors.
“This is the time to be building. There is so much to invent. There is so much to change.”
During a 2 ½-hour presentation assisted by executives from Expedia, Goldman Sachs and the National Football League, new launches included six services and capabilities for connected devices at the edge: AWS IoT 1-Click, AWS IoT Device Management, AWS IoT Device Defender, AWS IoT Analytics, Amazon FreeRTOS, and AWS Greengrass ML Inference.
Amazon FreeRTOS is an operating system that extends the functionality of AWS IoT to devices with very low computing power, such as lightbulbs, smoke detectors, and conveyor belts.
Meanwhile, AWS Greengrass ML Inference is a new capability for AWS Greengrass that allows machine learning models to be deployed directly to devices, where they can run machine learning inference to make decisions quickly, even when devices are not connected to the cloud.
“The explosive growth in the number and diversity of connected devices has led to equally explosive growth in the number and scale of IoT applications,” said Dirk Didascalou, vice president of IoT at AWS. “These new AWS IoT services will allow customers to simply and quickly operationalize, secure, and scale entire fleets of devices, and then act on the large volumes of data they generate with new analytics capabilities specifically designed for IoT.”
According to an AWS press release, when considering IoT, “many customers just want an easy way to get started by enabling devices to perform simple functions. These are functions like single-button devices that call technical support, reorder goods and services, or track asset locations.
“At scale, IoT solutions can grow to billions of connected devices. Today, this requires customers to spend time onboarding and organizing devices, and even more time integrating multiple systems to manage tasks like monitoring, security, auditing, and updates,” the company said.
Other new product launches and services included:
Alexa for Business: A service that provides employees with an intelligent assistant capable of tasks such as starting conference calls, controlling conference room equipment, scheduling meetings, keeping track of tasks and even reordering supplies.
Machine Learning Services: Amazon SageMaker, which allows developers and data scientists to build, train, deploy and manage their own machine learning models. Also introduced were the following application services: Amazon Transcribe for converting speech to text, Amazon Translate for translating text between languages, Amazon Comprehend for understanding natural language and Amazon Rekognition Video, a computer vision for analyzing video batches an in real-time. Accompanying the services is a hardware device called AWS DeepLens, a wireless video camera capable of running real-time computer vision models in order to give “developers hands-on experience with machine learning.”
New Container Capabilities: Announced was Amazon Elastic Container Service for Kubernetes and AWS Fargate that allows customers to launch and run containers without provisioning or managing servers or clusters.
New Database Capabilities: Announced were upgrades to Amazon Aurora and DynamoDB and the introduction of Neptune, a graph database service.
In his keynote speech, Jassy said that in the database space, the “last 20 years have been a very uncomfortable, unpleasant place for most companies with the database providers they have had to use. These are companies that are very expensive, that have high lock-in, are proprietary.
“Earlier this year, Oracle overnight doubled the price of their software to run on AWS and on Microsoft. Who does that to their customers? Somebody who doesn’t care about their customers. Somebody who uses customers as a means to their financial ends and it’s why customers are trying to move as fast as they can to the open engines.
“The landscape of how people use databases today is really different from what’s been the case over the last number of years. You don’t use relational databases for every application. That ship has sailed. Modern companies that use modern technology are not only going to use multiple types of databases and all their applications, but many are going to use multiple types of databases in a single application.”
The machine learning announcements certainly piqued the interest of George Papayiannis, co-founder and chief technology officer for Toronto-based Vena Solutions, a provider of cloud-based corporate performance management (CPM) software. Vena, which boasts it is the fastest growing cloud CPM company and the only one to embrace and not replace Microsoft Excel, provides budgeting, planning and revenue forecasting software for medium and large sized organizations.
“The biggest problem with machine learning and KPI (key performance indicators) is structuring the data. For us at Vena, the data is somewhat structured. It’s in OLAP – it’s in the multi-dimensional database.
“We have started to think about how we can use that data to provide insights into the data itself. Looking back at the historical data, what can we tell a company about their data that they didn’t know about. You have to first solve the idea of insight before you can even think about foresight.
“I have been thinking a lot about machine learning and AI models and how I can structure and think about OLAP data in the context of what a customer needs to know from their standpoint about their data.”
Papayiannis termed the machine learning announcements exciting in that they “provide me a benchmark now where I believe we will be able to use these services to get a baseline in terms of what we can expect out of machine learning and AI in a very fast, simple way. From there, we can then challenge ourselves to use those technologies and others to see what we can do better.”
As for the overall movement to the cloud, he added that over the last six months he has seen a “rapid, rapid cloud adoption from large enterprises that previously had showed some reluctance.
That sentiment aligns with results of a recent study from Toronto-based AWS partner Scalar Decisions Inc.
Findings revealed that almost three quarters of Canadian organizations have migrated at least some of their IT delivery to the cloud. That’s up from 59% in 2016. Scalar’s second annual Cloud Study shows Canadian organizations have moved past security concerns as a barrier to adoption, and are now focused on maximizing the return on their investment in cloud delivery.
According to the study, highly publicized security breaches have not caused organizations to avoid adopting cloud-based delivery.
While 35% of companies saw security as a factor in determining how much to use the cloud, 90% of respondents overcame those concerns to move forward with some level of cloud integration.
“The cloud has become not just a standard IT delivery model, but the standard IT delivery model,” said Rene Heroux, Scalar’s chief technology officer, who attended the conference. “While organizations have ongoing concerns about security, they now see the cloud as a potential solution to those risks, not as a barrier to adopting the technology. Businesses are now looking to reduce costs and increase the efficiency of their cloud infrastructure by developing new software and processes.”
The company added in a release that the next frontier for cloud technology is moving from achieving basic benefits such as right-sizing, capital retention, and scalability to more advanced features like continuous integration, continuous development, and continuous deployment process automation.
Results revealed that companies are planning to boost their spending on cloud services to reap those efficiencies. Organizations plan to increase the share of their IT budget devoted to the cloud from less than 20% to almost 30% in the next 36 months.
As expected, there were dozens of examples at AWS re: Invent illustrating cloud initiatives in action, but one of the most compelling comes not from a Wall Street financial giant or a sports league, but the hard working rice farmers in Asia and Africa.
In his keynote speech, Dr. Werner Vogels, chief technology officer at Amazon, referenced a Philippines-based organization called the International Rice Research Institute, which develops rice varieties designed to withstand “drought, flooding, disease and other potentially damaging events.”
Its target, said Vogels is to make the life of the poorest farmers in Asia and Africa that much better.
“One of the systems they have built is how to apply fertilizer at the right moment in these small patches of land that these farmers have. They built a digital system, but nobody wanted to access it because they didn’t have Web pages, they didn’t have computers. These farmers don’t have access to that.
“The big success came when they put a voice interface in it. Now, a farmer can go to the village phone, call a number, select from 27 different dialects, describe their patch of land in a sort of fuzzy way, machine learning goes off, comes back and describes exactly how much fertilizer to buy and when to apply it and the farmer has good advice. This has reduced the amount of fertilizer by 90% and doubled the crop yield.”
The big success of this overall system is putting a natural interface in it using voice and to that end, says Vogels, voice will be the next “major disruptor” in computing.