AWS re:Invent 2022: All the highlights and key announcements from this year's cloud computing conference
5 Things to know about Re:Invent
Hear from AWS CEO Adam Selipsky, Amazon CTO Dr. Werner Vogels, and more leaders at the Amazon Web Services (AWS) 11th re:Invent cloud conference, November 28 through December 2 in Las Vegas. Each year, re:Invent features leader keynotes, new service announcements, fun, and inspiration. Here are five things to know about this year’s re:Invent:
1. Expect a busy week of news and announcements
re:Invent is an opportunity to check out all the latest news and developments in compute, databases, analytics, machine learning, and storage. Follow all the key announcements, and get a peek at the newest cloud technologies from AWS at the Amazon Press Center.
2. re:Invent is in person with more programming
After going virtual in 2020, and offering a hybrid event in 2021, re:Invent is offering more in-person events so all attendees—from customers to partners to aspiring technologists—can learn from experts and each other. The team has designed an in-person program with nearly 2,300 sessions. AWS is live-streaming all keynotes and leadership sessions for virtual attendees, and all presentations, including breakout sessions, will be captioned and broadcasted. To catch the re:Invent action remotely, register to virtually attend.
3. Five keynotes and 22 leadership sessions are scheduled throughout the week
AWS leaders, including CEO Adam Selipsky, will host keynotes, announcing the latest product launches and sharing inspiring customer stories. Other keynote speakers include Amazon Chief Technology Officer Dr. Werner Vogels, Senior Vice President of AWS Utility Computing Peter DeSantis, Vice President of AWS Worldwide Channels and Alliances Ruba Borno, and Vice President of AWS Database, Analytics, and Machine Learning Swami Sivasubramanian. re:Invent will feature 22 leadership sessions. Check out the full agenda.
4. re:Invent is an opportunity to learn directly from AWS Cloud computing experts
At its heart, re:Invent is a learning conference, offering builder labs, bootcamps, gamified learning, and hundreds of technical sessions, from the introductory to the most advanced. Attendees get to dive deep with new technologies, and they can practice new ways of working and hone their skills alongside their cloud-community peers.
5. Expect plenty of interactivity with live demos and a robotic tap room
Participants will have opportunities to explore demos and interact with technology, including meeting a robot bar-keep, seeing a basketball free-throw analyzer, and playing a cloud-skills game. AWS is also hosting a showcase for sustainability, which highlights how technology is being used to address challenges like water conservation and decarbonizing operations. And the AWS Disaster Response rolling lab—a technology-packed truck—shows the benefits of cloud capabilities during disaster responses. Finally, attendees can catch the annual AWS DeepRacer League Championship, where machine learning meets model cars racing autonomously around a tough track.
AWS re:Invent 2022 is a go! AWS Principal Solutions Architect Rudy Chetty gives us a look at the action kicking off in Las Vegas, Nevada.
Senior Vice President of AWS Utility Computing Peter DeSantis gives audience a ‘peek under the hood’
In the re:Invent kickoff keynote Monday, November 28, in Las Vegas, Nevada, Senior Vice President of AWS Utility Computing Peter DeSantis detailed how the ongoing pursuit of greater performance at AWS never comes at the expense of security or cost. If that sounds like a tough combo to pull off, it is.
Whether it’s each new generation of AWS-designed Nitro systems (a combination of hardware and software, which we’ve been developing for the better part of a decade) that are more cost effective and energy efficient than the one before, or the latest computing instances that take full advantage of custom silicon chips, DeSantis gave example after example of how AWS is committed to long-term investment, and innovating from the ground up.
Performance in the AWS Cloud means offering customers an ever-increasing ability to ingest and analyze more and more data–simply, quickly, and, here’s the kicker, at lower cost. It’s a dance, DeSantis said, between hardware and software.
And performance isn’t only about what you can see, said DeSantis. It’s also about what you can’t. Nothing shouts performance more than a Formula 1 car, or what goes on behind the scenes to bring the fastest possible cars to the track.
On hand to drive home the point was Jock Clear, senior performance engineer for Scuderia Ferrari. All his team has to do, said Clear, is produce the fastest car over one lap–a straightforward mission that is anything but straightforward in how to make it happen. The fastest car in the first race isn’t the fastest car at the end of the season. The driving strategy for a race when you are in front of the pack and can’t see what’s coming up, or in second place and can clock the leader is different. It’s all about the tradeoffs, and how to trade smart. It’s here where innovation is key. “We can’t rely on the tried and trusted,” Clear said. “The competition is too tough.”
Clear talked about how together with AWS, Ferrari has developed an off-car, virtual ground speed sensor that uses AI and Machine Learning technology to provide engineers with more reliable data, faster, allowing it to save weight in a sport where every gram counts. Clear and team are also developing race strategies using AWS ML, as well as a new app, set to launch next year, which will give F1 fans a backstage pass (their own peek under the hood) with exclusive interviews with drivers and engineers.
From F1 trying to gather data to understand every aspect of a car’s performance, to weather forecasters trying to predict the next storm or healthcare companies developing treatments for disease, high performance computing and machine learning are two areas set to benefit the most from the continued innovations in some of the ‘fundamental aspects of the cloud’ DeSantis described. He closed by saying he’d only been able to cover a fraction of the investments AWS has been making, but that people should stay tuned this week.
There’s more to come.
What does it take to build the fastest F1 car on the track?
Jock Clear, senior performance engineer for Scuderia Ferrari, told the re:Invent audience what goes into the design process during the Monday Night Live keynote.
Take a look at the news from AWS CEO Adam Selipsky's keynote
Amazon Web Services (AWS) CEO Adam Selipsky made a host of service announcements during his re:Invent keynote this morning. Here’s an overview that includes links to press releases if you’d like to read in more detail.
The end of toilet paper shortages?
Whether it’s waiting seven months for a sofa to be delivered, arriving at your local grocery store to find they’ve run out of toilet paper, or not being able to buy the car you want because of a lack of microchips, it’s likely you’ve been affected by global supply chain issues one way or another these past few years. Getting us the goods we want, on time, and at low cost, has been harder and harder for businesses across all industries as they’ve had to deal with unprecedented volatility created by everything from natural disasters, to rapidly changing geopolitics, to resource shortages. Anticipating these kinds of events is no mean feat, especially when most companies rely on manual analysis of data in inconsistent formats, across multiple systems. This is where AWS Supply Chain promises to transform the status quo for shoppers and supply chain managers alike. The new cloud-based application builds on nearly 30 years of Amazon.com supply chain experience to give businesses in any industry a real-time, unified view of their supply chain data. It uses machine learning and other AWS technologies to provide insights and make recommendations on how to find trends more easily, and respond to potential risks faster. Read the press release to learn more.
Finding the right data to make the right decisions
In most large organizations, some people produce data, such as data scientists and engineers, and others use data to make business decisions, such as analysts. But decision makers can’t always find the data they need. And even if they do, they might not be able to get hold of it. Many companies rightly have a range of strong access controls and data governance policies in place to manage what can be—in the case of industries including insurance, automotive, and energy—petabytes (in other words, massive amounts) of information. And that information can live across hundreds of different sources, owned by multiple teams. Amazon DataZone, a new data management service from AWS, is designed to help broad groups of users easily find, organize, and share data across their organizations, use the best of AWS’s analytics tools to analyze it, and all the while, ensure it’s governed appropriately. The result? Businesses can drive more innovation by empowering people to make more timely, better-informed decisions, based on the most accurate and up-to-date information available. Read the press release to learn more.
Using AWS to unlock the value of your data
Global financial institution Goldman Sachs and jet engine component and system provider GE Aerospace are just two examples of AWS customers managing vast, complex data sets, and—due to that vastness and complexity—rely on multiple AWS database and analytic tools to extract value from it. Selipsky announced two new AWS capabilities that will make it easier for customers to connect and analyze data across data stores, without having to move that data across different services. Eliminating the need to extract, transform, and load (ETL) information in this way can help customers unlock even greater value from their data. Read the press release to learn more.
Making it easier to map megacities and much more with dynamic, large-scale simulations
City traffic patterns, crowd movement in a venue, and factory floor layouts are all examples of dynamic environments that can be digitally modeled in 3D to find out how they will respond to expected and unexpected changes, without actually having to test these scenarios in real life. If, say, you’re responsible for urban planning in a city, you can create a spatial simulation to investigate the impact of different road closure combinations. It’s an extremely useful way of visualizing and predicting potential outcomes, as well as creating immersive training spaces. But the larger the scale of what you’re trying to model—from mapping populations in megacities to tracking international logistics operations—and the more nuanced the problem you’re trying to solve, the more complicated it gets and the more compute resources it requires. Scaling a spatial simulation with millions of interacting objects has often required years of investment, specialized hardware, and a team of high-performance computing experts. That could all be about to change with the launch of AWS SimSpace Weaver, a new service that allows customers to build, operate, and run large-scale spatial simulations without having to worry about the capacity of their own hardware, or rely on specialist skills. Read the press release to learn more.
Analyzing data across multiple organizations while protecting sensitive information
Whether it’s planning advertising campaigns, making investment decisions, or analyzing clinical research, organizations across industries are increasingly looking to collaborate with others on their collective datasets, but without sharing or revealing underlying information. Take advertising: brands, media publishers and their partners rely heavily on data to gain insights that improve the relevance of their campaigns for consumers. But that data is often spread across different channels and applications. And, to protect sensitive consumer information, one company will need to provide a copy of their user-level data to another, and rely on contractual agreements to prevent misuse. "Data clean rooms" help solve this challenge. They allow multiple parties to combine and analyze data in a protected environment, without sharing raw data. The problem is that they’re hard to build, requiring complex privacy controls and specialist tools. AWS Clean Rooms is a new service that allows customers to build a data clean room in minutes and collaborate with any other company in the AWS Cloud to generate unique insights about their data, while protecting sensitive information. Read the press release to learn more.
Enabling faster threat detection, investigation, and incident response
Security is a top priority for any business, whether that’s protecting data, uncovering policy violations or preventing fraudulent activity. Most companies rely on log and event data from different sources to identify potential threats and vulnerabilities, assess security alerts, and respond accordingly—but they must first aggregate it all and put it in a consistent format in order to analyze it. This takes time and can slow down a team’s ability to detect and then respond to issues. With AWS's new Amazon Security Lake service, customers can automatically create a security data lake in their AWS account with just a few clicks. It means organizations can automate data collection from a variety of different sources, and collect, store and manage it more easily, to enable faster threat detection, investigation, and response. Read the press release to learn more.
Using AWS-designed silicon to solve complex academic, scientific and business problems
Most people know AWS as a company that offers cloud computing services, not one that designs its own chips. In fact, we have our own family of chips and accelerators, with each new generation building on, and improving on, what came before. All of our chips are designed by our Annapurna Labs team, based in multiple locations around the world. Their work is featured in both Peter Desantis’ and Adam Selipsky’s keynotes, with the announcements of three new Amazon Elastic Compute Cloud (Amazon EC2) instances—virtual ‘machines’ that run workloads in the cloud—powered by AWS-designed chips. Among other capabilities, the new instances are built to offer the best price performance for running high performance computing workloads at scale on Amazon EC2, helping customers solve some of their most complex academic, scientific, and business problems—from genomics processing to mathematical modeling to weather forecasting. Read the press release to learn more.
Integrating more intelligent business intelligence
Paginated reports (reports that are designed to be printed or shared, because they’re formatted to fit well on a page) have for decades been the standard format of reporting across industries. And while they’re still used to share important operational information, such as daily transactions or weekly business updates, it’s generally not been possible for companies to integrate them with modern business intelligence dashboards. This means they often use multiple systems—one for business intelligence reporting and one for paginated reports—which can be costly and inefficient. Amazon QuickSight Paginated Reports is a new capability that will make it easy for customers across industries—from hospitality to industrial manufacturing, to ed-tech—to create and share business critical reports and data across their organizations. It’s one of five new capabilities for Amazon QuickSight (AWS’s serverless business intelligence service built for the cloud) announced by Selipsky today in his keynote. Read the press release to learn more.
Ukraine minister on how technology is helping his people plan for the future
Government doesn’t need to digitize bureaucracy, said Mykhailo Fedorov, Ukraine’s Minister of Digital Transformation. Government needs to simply fix people’s problems, fast.
Fedorov is focused on the future. As a technologist that is always part of the job, but in his ministerial role, he’s looking ahead with a sense of urgency and a massive dose of courage. For all the horrors he and his fellow Ukrainians are experiencing in the present, Federov communicated a sense of optimism to the audience who gathered at re:Invent to hear him speak on Tuesday.
Describing the war as something out of science fiction, he said that technology has helped Ukraine fight, survive, and plan for a time when the conflict ends.
Federov related how an early Russian attack destroyed the building in which the Ukrainian government stored all its backup data—the backup records of its citizenry and its institutions, the data that literally described an entire nation.
In the aftermath, the Ukrainian government quickly passed legislation that allowed it to move state registries and data to AWS. Using AWS Snow Devices brought into the country, the government was able to move vital information to the cloud. To date, 38 government authorities and 24 universities have moved their data, and the services they offer to Ukrainians, onto AWS, Fedorov said. What that means is that education continues, the rightful ownership of property is maintained, as is—sadly—a record of what has been damaged and destroyed.
While cruise missiles have rained down on cities across Ukraine, Federov and his team have been huddled around laptops building systems and services that are helping people access banking, healthcare, school, even job training for careers in technology. When information has been cut off, they have built digital television and radio channels to get the official Ukrainian word out.
Along the way, Ukraine has created a digital government infrastructure that the rest of the world may do well to emulate.
“For nine months, Ukraine has been fighting for freedom and democratic choice,” Federov said. “In many ways, this is a war of technologists. Me and my team believe in big data, analytics, and constant improvement, but as technologists in Ukraine, we are using technology (on the) side of light.”
“We want to build the world’s most convenient state in terms of public services,” he said. Then, half joking, he made an appeal to the audience to send any services that might be useful his way. “Ukraine is now the best testing ground for your product,” he said.
All Builders Welcome grantee finds new community at Re:Invent
There are more than 50,000 attendees at re:Invent this week. But few, if any, have a journey quite like that of 28-year-old software engineer Adesola Adesina.
Adesina was pursuing a master’s degree in sociology five years ago when she knew she needed to make a change. “I didn’t like what I had built for myself,” she said.
She moved from Lagos, Nigeria, to Washington D.C., at the age of 10, with a dream of becoming an attorney like her idol Clair Huxtable, from the popular sitcom The Cosby Show. Growing up, she didn’t pursue engineering because she considered herself “bad at math.” But in 2017, she found herself camped out in a sleeping bag on her mom’s floor after a break-up, and realized she needed to drop out of grad school to pursue a career that would provide more financial stability—and more personal satisfaction.
She began a reinvention. She signed up for free trainings at a Python 101 bootcamp. She got a new job as a research analyst. She dug into online instructional videos. Adesina continued a path of exploration, zig-zagging through different roles and learning opportunities until she saw a type of software engineering that she connected with. “I finally found myself in a place where I could see engineers doing what engineers do, and I said, ‘I want to do that.’”
This week, Adesina arrived at re:Invent as an All Builders Welcome grantee, provided with travel and access to the conference, mentorship, and training sessions. AWS created the All Builders Welcome grant program to support a future of tech that is diverse, inclusive, and accessible. It’s open to early-career technologists from underrepresented backgrounds and provides opportunities to learn, network, and build.
Adesina is now a product solutions engineer at a supply chain logistics startup. She credits members of organizations like Women Who Code and Black Code Collective for coaching and guiding her along the way, but she knows she still has much to learn.
“When you’re self-taught, there are a lot of giant gaps in your knowledge,” Adesina said. She wanted to come to re:Invent to bridge some of them. “A lot of our architecture is cloud native. It’s a type of thinking I had never been exposed to. I’m realizing it’s a whole new world I can sink my teeth into.”
Photos worth 1,000 terabytes*
It's not all interactive tech demos and robotic tap rooms here at re:Invent in Las Vegas. A stroll to the second floor of the Venetian offers a calmer, unexpected perspective on the cloud. This gallery of 20 large-scale photographs celebrates individuals and communities connected to AWS, and not necessarily the ones you might imagine—from farmers to fishermen, fencing contractors to middle school students, even a prime minister. It's a different take on the people and places that make our technology tick. (*And yes, that’s a petabyte.)
AWS Free Throw Analyzer vs. the ‘granny shot’
We’re all guilty of scoffing at a missed free throw from the safety of our screens. But when challenged to sink a shot, many of us quickly realize how difficult it can be.
At re:Invent, the bravest attendees took their pride to the foul line, putting their basketball skills to the test. The AWS Free Throw Analyzer invites contestants to take five shots, which are monitored by a camera that is connected to an AWS Snowball Edge Device. A machine learning algorithm dissects each shot, providing real-time data on the speed, angle, trajectory of the ball, and position of the player.
We set out to see if we could hack the system with the tried-and-true "granny shot" which, despite being a ridiculous way to shoot a ball, has been scientifically proven to be the most accurate free throw technique.
Charting a future of carbon neutrality through predictive energy data
Global utility company ENGIE has a 180-year legacy that started during the industrial revolution. The company, based in France, is now charting its own digital revolution as it aims to accelerate the transition to a carbon-neutral economy.
For ENGIE, becoming truly digital means being able to aggregate, store, and analyze vast amounts of data—both from its own sources, and also from third parties, like weather data repositories. Being able to compile and analyze information from many places (and store it using services like Amazon S3) is key, said Biljana Kaitović, who heads digital and IT for the company and spoke at re:Invent during AWS CEO Adam Selipsky’s keynote Tuesday.
“The fact that we can bring so much data from different places—and connect through APIs in a way that we couldn’t before—is extremely impactful for all companies that are not digitally native,” said Kaitović when we talked to her later that day. “Most older companies have a huge legacy, with disparate systems and data not talking to each other. It’s a real game changer to be able to take enormous amounts of data and actually start making sense of it.”
One of ENGIE’s notable use cases ties directly to its net-zero carbon goals. Predicting the production and availability of renewable energy sources like wind, solar, and hydro, using weather data and operations data, helps to manage their intermittency. Also, data analytics, machine learning, and IoT help to set optimal maintenance timing. In other words, predicting equipment failure before it happens. This in turn drives efficiency and reduces downtime, saving energy, and making it more available for customers when they need it.
“Overall, digitalization is something that I believe is necessary for survival of industrial companies in tomorrow’s world,” Kaitović said. “Unless an enterprise becomes truly digital, I’m not sure it’s going to survive in the long term."
Data, data, and more data
The "vast and challenging" realm of data and the services AWS provides to help customers mine, manage, and stich together data from different sources to draw new insights from it was a main theme of AWS CEO Adam Selipsky's keynote yesterday. If you missed it live, don't worry. It's now available to watch on demand.
Don’t let your data lake become a data swamp, says Vice President of AWS Database, Analytics, and Machine Learning Swami Sivasubramanian
Helping people better collaborate on data and derive valuable insights from it was at the heart of yesterday’s new AWS service announcements. And that message came through loud and clear again this morning from Vice President of AWS Database, Analytics, and Machine Learning Swami Sivasubramanian. Quality tools that endure over time and quality data are the key ingredients for long term growth, Sivasubramanian pointed out in his re:Invent keynote on Wednesday. Not to mention a secure system of shared governance for disconnected teams and disconnected data stores. Sivasubramanian announced more new AWS offerings relating to database, analytics, and machine learning. Here’s an overview of the main news, with links to press releases if you’d like to read in more detail:
Keeping your data updated
All supermarket managers know the importance of keeping their produce fresh, but what about their data? If you’re in charge of a retail chain and your team is analyzing data that’s not as up-to-date as it could be, you might underestimate demand for a product—leading to shortages, and missed revenue. Helping customers monitor, measure, and manage the quality of data in their data lakes is just one of five new capabilities across AWS’s database and analytics portfolios announced by Sivasubramanian. The end goal? To continue to make it easier for customers of all sizes, across all industries to maximize the value of their data to derive new insights and make faster, data-driven decisions. Read the press release to learn more.
New ways of using machine learning for climate science, disaster response, agriculture, and more
From preventing wildfires, predicting what passengers want to eat on flights across Europe, protecting sperm whale populations, and helping farmers grow more of a traditional ingredient, machine learning (ML) is fast becoming an integral part of our day-to-day lives, and the range of problems it can be put to work on is endless. But only a few years ago, building, training and deploying ML models was a painstaking task, and one only a few highly specialized experts within an organization had the ability to work on. AWS launched Amazon SageMaker in 2017 with the aim of opening up ML to more people, and making the process of building ML models easier, faster, and more cost effective. Our commitment to democratizing ML in this way has continued ever since, with Sivasubramanian announcing eight new capabilities for Amazon SageMaker during his re:Invent keynote. One of these includes a new level of support for geospatial data, which will make it easier to develop models for climate science, urban planning, disaster response, precision agriculture, and more. Read the press release to learn more about all eight new Amazon Sagemaker capabilities and how they will make it even easier for customers to take advantage of ML at scale.
How the cloud is improving employee safety and delivering cleaner energy
Some jobs are clearly more high risk than others. Working on a hydroelectric dam, for example, requires you to pay infinitely more attention to your surroundings than sitting at your desk tapping away on a laptop. But what if you could get a helping hand when looking out for potential hazards? Not to mention an extra pair of eyes, and ears? That’s the case for employees of Brookfield Asset Management—one of the world’s leading investors in renewable power—at three of its hydroelectric dam sites in the U.S., who must be onsite to ensure the facilities are running properly.
All employees working at these locations wear what’s known as Blackline Safety’s G7 Lone Worker technology—cloud-connected IoT devices that continuously, and at high-speed, stream location and other data directly to the Blackline Cloud, which runs on AWS. The devices automatically detect if an employee needs health or safety assistance and send an alert to a monitoring team who can immediately dispatch help.
Advancing workplace safety isn’t the only way AWS and Brookfield are collaborating. AWS is helping Brookfield modernize its operational systems, giving it the scalability, reliability, and innovation it needs to deliver clean energy around the world. Meanwhile, Brookfield will provide 601.6 megawatts (MW) of clean wind and solar energy capacity to power Amazon’s operations in Europe, North America, and India. In total, these projects are expected to generate enough energy to power more than 120,000 U.S. homes per year. To learn more about the AWS and Brookfield collaboration, announced today during re:Invent, read the press release.
10 things we learned about AWS customers at re:Invent yesterday
Migrating to the cloud is all about digital transformation and its effects on business outcomes and bottom lines. But it’s hearing how customers—from startups to long-established enterprises—are actually using all this breakthrough technology that really brings it to life.
Here are 10 things we learned about AWS customers from AWS CEO Adam Selipsky’s keynote Tuesday:
- Germany’s BMW Group is “revolutionizing” the driving experience with its connected-car platform.
- Video game maker Riot Games is processing approximately 500,000 events per second to continually improve its games.
- Epic Games’ Fortnite video game supports hundreds of millions of players, with up to 13 million players interacting together in real time.
- As the war with Russia unfolded, PrivatBank, Ukraine’s largest bank that had served 40 percent of its population, moved all of its operations to the AWS cloud–330 IT systems and four petabytes of customer data–in just 45 days to safeguard customers’ abilities to continue banking.
- There are more than 1,000 so-called “unicorn” startups globally, according to PitchBook, and 83 percent of them run on AWS. And more than 90 percent of the startups on the Cloud 100 list run their businesses on AWS. Unicorn startups on AWS include luxury fashion e-tailer Ssense, cloud security platform provider Wiz and DataRobot, an artificial intelligence (AI) cloud company.
- Travel company Expedia Group processes more than 600 billion AI predictions per year, powered by 70 petabytes of data.
- Pinterest, an image-sharing and social media platform, stores an exabyte–a million terabytes–of data on Amazon S3, AWS’ object storage service.
- The Options Clearing Corp., which serves as the central clearing warehouse for all listed equity options traded in the United States–part of the “central nervous system” of the financial markets–will be moving its core workloads to the cloud, thanks to a recent move by the U.S. Securities and Exchange Commission. It’s a once-in-a-generation technology decision, according to Selispky, who told the audience how excited he was that OCC is running its “mission-critical” applications on AWS.
- Formula 1's racing car design requires simulations involving more than 550 million data points to model the aerodynamic weight of their cars. On AWS, Formula 1 was able to reduce the simulation runtime by 70 percent, from 60 hours down to 12, creating a new car with less than half of the turbulent airflow.
- The Nielsen Company, a measurement and data analytics company, uses AWS to process hundreds of billions of advertising-measurement events every day, and to ensure it can scale its infrastructure with the right compute to support this demand.
The letter that sparked a new tech education program for underrepresented communities
When Houston Community College (HCC) professor Dr. Raymond Brown wrote to the AWS Machine Learning University (MLU) to share how he'd been using its resources to open up artificial intelligence and machine learning to his students, he had no idea his letter would be the catalyst for a new educational program.
AWS MLU was originally conceived to provide individual users with self-service machine learning training from Amazon's own scientists. When Brown contacted the team, it was the first time they’d heard about an educator taking the content and adapting it for teaching—and it gave them an idea.
Inspired by Brown’s resourcefulness, AWS MLU built a new, free program—announced today by Vice President of AWS Database, Analytics, and Machine Learning Swami Sivasubramanian during his re:Invent keynote—to help community colleges, minority-serving institutions (MSIs), and historically Black colleges and universities (HBCUs) teach database, artificial intelligence, and machine learning concepts to their students.
The program, which includes an educator enablement virtual bootcamp to introduce teachers to the content, is expressly designed to address opportunity gaps faced by students who are historically underserved and underrepresented in technology disciplines. It is available to all U.S. colleges and universities, with priority consideration given to community colleges, MSIs, and HBCUs for the educator-enablement component.
Find out how MLU content helped Dr. Brown launch an AI program at HCC that should see it become the first community college in the U.S. to offer a bachelor’s degree in AI in fall 2023.
It’s all about the cells
Lyell is pioneering novel T cell reprogramming technologies designed to breakdown major barriers in the fight against solid tumor cancer. Here’s how they’re using cloud technologies from AWS.
Tackling Earth’s biggest problems with out-of-this-world data analysis
What does mining exploration have in common with your favorite sustainable cosmetics? How does predicting next year’s soybean crop relate to preventing landslides?
The answer? They all rely on the power of geospatial data.
Descartes Labs, a geospatial intelligence company founded out of Los Alamos National Laboratory in, New Mexico is collecting and analyzing this data—gathered from satellites, aircrafts, and other sources—for use across a wide range of industries. Their aim is to tackle some of Earth’s biggest challenges in the process.
“Data is everything to Descartes Labs,” said the company’s Head of Commercial Deanna Lyn when we spoke to her at re:Invent Wednesday. “We bring petabytes (of information) from disparate sources—everything from Earth observation data, automatic identification system (AIS) data, commodity trading data—and we bring it into a place that allows companies to quickly blend that data together and find answers to hard questions.”
Thanks to Descartes Labs, customers in the mining industry can now identify the best places to start digging—saving months of manual exploration time and preventing environmental hazards like landslides. Consumer goods companies like Unilever can monitor the carbon footprint of their palm oil supply chain, which ensures cosmetics and other products are environmentally sustainable. And commodity traders can now predict the future growth of specific crops.
The Descartes Labs team relies on a host of AWS services and products to maintain and leverage its 20 petabyte data library. From Amazon S3 to optimize on data storage costs, to Amazon OpenSearch Service, a search and analytics suite that allows it to find veritable needles in haystacks by locating the precise pixels it needs to solve particular problems. The company announced this week at re:Invent that it’s migrating all core information technology infrastructure, including geoprocessing and analytics platforms, to AWS.
According to Lyn, this will allow for more innovation in machine learning and predictive modeling in future, while the team will remain focused on addressing everything from climate change to food security.
Data beats intuition
How to extract value from data is a theme that’s come up again and again here at re:Invent. Why? Because it’s a challenge for organizations of all sizes, across all industries. Most organizations lack centralized repositories and data often is fragmented—inconsistently collected, stored, and controlled across many business units. Remedying that requires complex pipelines to move data to the right place, and mechanisms for the right individuals to access it, when and where they need it. Vice President of AWS Database, Analytics, and Machine Learning Swami Sivasubramanian had some answers for this during his re:Invent keynote Wednesday. He delivered a blueprint for a strong data strategy, at the center of which were three main elements:
- A future-proof data foundation—one that can keep up with ever-growing volumes of data by performing at a high scale. It should remove the burden for an organization, so teams can spend less time managing and preparing data and more time getting value from it. And it should have the highest level of reliability and security, in order to protect that data.
- Solutions that weave a "connective tissue" across organizations. Connecting data across an organization requires solutions from automated data pathways to data governance tools, Sivasubramanian said. “Not only should this connective tissue integrate your data, but it should also integrate your organization's departments, teams and individuals.” And data governance enables “safe passage” for disconnected teams and disconnected data stores, so organizations can collaborate and act on their data, he said.
- The right tools and education to help democratize an organization’s data. "Democratizing data" as Sivasubramanian put it, can spark innovation across an organization—empowering everyone from interns to product managers to business analysts with no technical expertise. From no-code tools that enable non-technical employees to do more with data, to training programs that help people develop technical skills in critical areas such as machine learning, you can help your employees (and future employees) organize, analyze, visualize and derive insights from data, and cast a wider net for your innovation.
“I strongly believe data is the genesis for modern invention,” Sivasubramanian said. “To produce new ideas with our data, we need to build a dynamic data strategy that leads to new customer experiences as its final outcome.”
Getting applications up and running in minutes
Creating a new application that takes advantage of the latest cloud technologies requires software development teams to seamlessly work across a range of tools and environments to build, test, and finally deploy it. Amazon CodeCatalyst, a new service announced by Amazon CTO Werner Vogels during his re:Invent keynote today, brings together the tools software development teams need to build and deliver applications on AWS more easily, and quickly. With Amazon CodeCatalyst, developers can bring their own code or select from a range of blueprints such as "web application" or "data pipeline," to get a new application development project up and running in minutes. Read the AWS news blog to find out more.
Builders put on a show at the re:Invent Builders’ Fair
There was axe-throwing analysis (plastic safety axes, thank you), a robot yoga coach, a recorder that played itself (the 2nd-grade plastic flute version), automated rock paper scissors, and a cloud-assisted beer brewing.
The Builders’ Fair in the Expo at re:Invent showcased the many ways cloud technologies—from image recognition to natural language processing and every flavor of database—can be put to work on almost anything.
What the AWS teams here have built illustrates the power and flexibility of cloud services, and in playful ways (engineers have fun too, after all). Two projects in particular—Kitchen Guard and American Sign Language Interpreter (ASL) 2.0—turned the cloud toward solving issues of accessibility.
Kitchen Guard: Helping visually impaired chefs prepare food that’s safe to eat
Is my beef cooked? It’s an important question for anyone grilling on the backyard barbecue. For colorblind and visually impaired people, it can be an especially tough one to answer.
Four AWS employees who originally met online as part of the same AWS Tech U virtual cohort set out to solve this problem nearly two years ago. This week at re:Invent, Gabrielle Dompreh (London), Lydia Khalfoun (Paris), Kimessha Paupamah (Johannesburg), and Saubia Khan (Dubai, United Arab Emirates) were finally able to meet in person for the first time and share their prototyped solution—an AWS-powered app called Kitchen Guard.
Their proof of concept allows anyone to snap a photo of a piece of food and then ask a question verbally. For example: Is my apple fresh? Is my cheese moldy? Is my steak raw?
The team used speech and image recognition tools like Amazon Lex and Amazon Rekognition to identify various foods, and then applied models to determine the food’s status. This week, they’ve been speaking to customers and colleagues about future applications for their tool, beyond assisting visually impaired customers—from food prep training to commercial kitchen safety for a broad market.
“I’m overwhelmed by the response,” said Khan, as she demonstrated the app alongside her teammates at the AWS Builders’ Fair in the re:Invent Expo Hall. “People are really interested in the different use cases.”
Converting American Sign Language into text and speech
AWS is well known for working backwards from customers when it comes to creating new products, but it was an intriguing conversation that served as development inspiration for AWS Solutions Architect Ashwini Rudra.
Rudra’s American Sign Language 2.0, a prototype featured at the Builders’ Fair, converts ASL fingerspelling into text and speech using a camera and multiple AWS Cloud services, including machine learning.
Rudra and his girlfriend were out for coffee one day when they saw a couple having a conversation in ASL, which neither he nor his girlfriend understood.
“I wanted to know what they were talking about. Were they fighting, or were they were expressing affection?” Rudra said.
He adopted his girlfriend’s suggestion of using a camera to read ASL, and his nascent machine determines the degrees of confidence to which each signed letter of a word is accurately read. The ASL Interpreter 2.0 uses Amazon SageMaker for machine learning, Amazon Simple Storage Service (Amazon S3) for object storage, and AWS IoT Greengrass, an Internet of Things, open-source, edge runtime and cloud service for building and managing IoT applications on devices.
“The possibilities are limitless,” he said.
NFL asks data pros to pitch in to prevent future injuries on the field
The National Football League (NFL) has been making strides in recent years by using AWS infrastructure and predictive modeling tools to prevent players’ injuries. Now, the league is banking on help from developers worldwide to help take player analysis to the next level.
The league announced its new NFL Contact Detection Challenge at a re:Invent panel on Thursday featuring former Arizona Cardinals wide receiver Larry Fitzgerald and Amazon leaders who work closely with sports organizations. Fitzgerald said the NFL’s data-driven injury-prevention efforts have allowed athletes to train more effectively and play more safely.
One of the biggest results so far is a 25% reduction in concussions among players—sustained over the past four years.
"Quantifying the risk of injury that players face in every possible in-game scenario is a crucial step in understanding how we can reduce that risk and ultimately prevent injuries," said Jennifer Langton, senior vice president of Health and Safety Innovation at the NFL, who also spoke on the panel.
The league’s newly announced challenge, in partnership with AWS, invites experts to design new ways to measure and analyze the timing, duration, and frequency of player contact during NFL games using artificial intelligence and related disciplines. This information will allow the league to better predict—and prevent—player injuries.
Visit the website to learn more about the challenge, which starts Monday and offers monetary awards for top finishers.
Navigating re:Invent for the first time
Flikshop founder Marcus Bullock took to the stage and to the session rooms for his first re:Invent this week. Flikshop is a Washington D.C.-based startup, built on AWS, that lets users send photos as postcards to incarcerated family and friends. Here’s three things that struck Bullock about the event:
1. At check-in and walking through the halls on my first day, what stood out immediately to me was: I am not the only Black person here. That was massive for me.
2. My struggles as a startup are the same in some ways as much larger organizations. It was a relief to know that I am in good company. I was up there on stage with Code.org and AARP—these giants—and I was fascinated to learn that they’re having some of the same challenges, whether around hiring, professional development, or introducing a new product, that I am.
3. I’m a non-technical founder building a tech company, and the conversations around technology were so thoughtful and in-depth, without all of the jargon you often get. I loved learning about new AI frameworks that we can think about using to set up flywheels that will help us scale much quicker. I don’t want to get stuck in the vernacular, I just want the technology to get to work for us, to help us leverage our data better and to reduce recidivism.
The connected brewery that’s tapping the cloud for insights
Who are AWS Partners? You probably know them already. From and CrowdStrike to Trend Micro and VMware, they’re companies that sell and support AWS enterprise products and services, and build their own solutions and devices around them.
Ruba Borno, Vice President of AWS Worldwide Channels and Alliances, demonstrated how AWS Partners play an integral role in helping customers innovate during her re:Invent keynote Wednesday.
One of these examples was Netherlands-based Schuberg Philis, which recently helped Heineken (the nearly 160-year-old brewing company) rethink its supply chain by connecting 16 production lines and 140 breweries across the globe to a new cloud-based platform. Heineken looked to new digital technologies to derive improved insights using data from its operations.
Working with AWS, Philis created an IoT platform that collects raw machine data from Heineken production lines and normalizes it so Heineken teams can analyze it to optimize production processes.
While connecting a global network of breweries that all ran different machines, with different production lines, and various suppliers with their own protocols, wasn’t without its challenges, it’s been hugely successful. That’s according to Michiel Maagd, Heineken’s global connected brewery leader, who says it’s also helping the company develop its next generation of employees.
“The platform helps us keep ownership of our data,” he said. “It avoids (creating) siloed insights. Most important, it helps us to easily scale use cases from one brewery to another brewery. Now our shop-floor workers have real-time insights into the production processes. They can see the temperatures, the speed of the lines, and we can also scale more clever use cases—like algorithms that help to define root causes of minor stops and slowdowns of the lines.”
“Our partners are critical to our customers’ transformation, and we believe that a better partner experience will result in a better customer experience,” said Borno. You can watch her keynote here.
Keeping your electric vehicle road-ready
If your electric vehicle (EV) is running low on juice, you need to know that a wall charger is ready (and working) when you drive up.
Availability of power points is just one of the things that Wallbox, a Barcelona-based EV charging and energy management company, wants to make sure its customers can rely on. With customers in 113 countries, Wallbox relies on AWS to make sure its technology is efficient and nimble. At re:Invent, the company announced it has migrated its entire information technology (IT) infrastructure, including design and manufacturing platforms, device and grid-management systems, and customer-facing applications, to AWS.
“EVs are rapidly changing. Charging is changing, and the supply chain is changing,” Brian Hall, Wallbox’s marketing director for North America, told us. “You have all these dynamic forces, and AWS certainly helps us to be more agile, more adaptive.”
Wallbox uses Amazon Kinesis, a service that collects, processes, and analyzes real-time streaming data, to collect data and monitor the status of its public and residential EV chargers around the world. Pulsar Plus, its residential charger, uses Amazon EC2 Spot Instances and AWS Graviton2 to power its software development, reducing overall IT costs by 70%.
The data Wallbox is collecting and analyzing using AWS is helping the company gain insights to optimize charger settings—making sure its EV-driving customers can get back on the road, fully charged, as quickly as possible.
Amazon CTO Dr. Werner Vogels wants you to use simulation
What was once the domain of a specialized few, special simulation is now available to everyone. You can experiment and better plan everything from traffic patterns to supply chain bottlenecks and forest regrowth.
AWS is bringing spatial simulation to every industry, allowing customers to more easily simulate everything from traffic patterns to public transportation networks to supply chain infrastructure. AWS SimSpace Weaver, now available, is a new, fully managed compute service that helps users run large-scale spatial simulations in the cloud without managing complex infrastructure.
“The real world around us is enormously complex,” Werner Vogels, vice president and chief technology officer at Amazon, said in his re:Invent keynote on Thursday. “There are thousands of events, and there are thousands of entities continuously around us. Think about … a big football game going on and a pop concert not far away from that. How does traffic go through the city to reach them? How do you manipulate that? What do we have to do to make all of that safe? If you would simulate that, you literally would have hundreds of thousands of different entities in there … position, velocity, behavior, interaction behavior.”
Those types of simulations used to be run on a single machine, because they require a significant amount of memory,” Vogels said. “AWS SimSpace Weaver makes spatial simulation accessible for everyone. It really lowers the barrier to it.”
According to Vogels, what makes AWS SimSpace Weaver unique is how the simulation is distributed. It allows users to break down a simulation into a grid—smaller, discrete spatial areas—and it distributes the task of running the simulation code across multiple Amazon Elastic Compute Cloud (Amazon EC2) instances.
“The simulation logic operates on each of the different cells in the grid,” said Vogels. “Each cell tracks the logic for that area and for all the entities in that cell. Every object is a separate entity … and the cells are distributed by the cluster of compute instances. The instances work together to process the entire simulation, as if you have one very large memory model, and it appears as a single integrated space with everything in it.”
Simulation plays a crucial role in innovation, because users can experiment with “what-if” scenarios that can’t be done in real life, according to Vogels. “You can experiment as much as you want without having actual impact on the real world before you take your systems to the real world,” he said.