It’s 8:00 AM on a Tuesday in Las Vegas, and you’re already feeling the adrenaline while the band plays Sweet Child of Mine as Adam Selipsky, the CEO of AWS, takes the stage. You heard that you needed to be in line by 7:00 if you wanted a seat, and you made it. You look around and realize that probably only 10% of the 50,000 in-person re:Invent attendees made it into this massive ballroom. This many people in person still feels surreal.
Selipsky did not disappoint; he delivered an inspiring keynote ranging from sustainability to historical inventions and customer successes to new AWS capabilities customers have been asking for.
The Issue of Our Generation
He believes sustainability is “the issue of our generation” and boasts Amazon as the world’s largest corporate purchaser of renewable energy. He announced the plan to power their operation with 100% renewable energy by 2025 (85% of the way there now); the crowd cheered. His announcement about AWS becoming water positive was especially welcomed by me as a Las Vegas resident. This means Amazon is committed to returning more water than is used by 2030.
How Machine Learning (ML) on AWS is supporting renewable energy
ENGIE is a large 180-year-old global energy company in France. Biljana Kaitovic (EVP – IT and Digital and group CIO) stated the annual global energy consumption could power Las Vegas for 5,000 years, which is projected to increase by 50% by 2050. ENGIE uses data and analytics to optimize renewable energy (sun, wind) using machine learning (ML) and IoT for predictive analytics for items like equipment failure. They have a common data hub storing over a petabyte of data on AWS S3, which is used across 1,000 projects. They use Amazon Redshift, Kinesis Data Streams, Glue, Athena, and SageMaker to train ML models. They have reduced cloud costs by 60% by scaling up and down based on working hours.
AWS and ENGIE partnership – Help Us Help You Help Us….
ENGIE provides green electricity to AWS from wind farms in the UK, US, Italy, and France.
“If you’re looking to tighten your belt, the cloud is the place to do it.”
In times of economic uncertainty, it can be tempting to cut back, but moving to the cloud can actually be more cost-effective.
- Carrier has reduced costs by 40%
- Gilead has realized $60M in savings over five years through cloud initiatives
- The ability to scale up and down helped Airbnb reduce their cloud costs when their business was impacted during the pandemic
Serverless, Serverless, and More Serverless
Selipsky spotlighted Aurora’s performance and cost-effectiveness; performance 5x that of MySQL, AND a tenth of the cost of commercial databases. AWS has a broad selection of serverless analytics services that can scale, and when I saw them span the big screens, it was quite impressive (Athena, EMR, Kinesis, Redshift, Glue, QuickSight, and Open Search is now serverless). Other attendees must have been just as impressed, as they were pulling out their phones and taking pictures of the screens.
Everyone’s Doing It; Even The Stock Market
Did you know that 83% of start-ups run their workloads on AWS? PrivatBank, the largest bank in Ukraine, moved all operations to AWS in 45 days. Nasdaq is on track to finish migrating the North American options market to AWS by the end of this year. Cloud is now the default platform for SEC-regulated workloads (we’ll talk about the OCC in a minute).
Who Loves ETL?
AWS integration between services seeks to eliminate the “thankless, unsustainable black hole of ETL.” Selipsky described query capabilities in Redshift and Athena running across data stores, including 3rd party applications. When he announced a zero-ETL integration with Aurora and Redshift, the crowd cheered again! Who doesn’t love the sound of a “zero ETL future?” Also announced was a Redshift integration for Apache Spark. No more building custom data pipelines between Aurora and Redshift or manually moving data around!
Lots of cool launches in data
Amazon DataZone has been launched. DataZone helps catalog, discover, share, and govern data. Analyze data across an organization with integrations between DataZone and Redshift, Athena, and QuickSight. Integrate Snowflake and Tableau through APIs.
Print-friendly reports were launched in QuickSight as well as ML-powered forecasting with Q. Business users can forecast sales for the next 12 months or even ask why sales went down without waiting weeks for the data team to run analyses.
Who has a good tagline for feeling even more secure?
Everyone knows Maslow’s hierarchy of needs; we need to feel safe. Selipsky was spot on when he described how feeling secure can either advance or limit us from exploring and building further technology. He used the stage to give the audience warm fuzzies about AWS security with a testimonial on trading and markets. Options Clearing Corp. (OCC) is the central clearinghouse for all US equity options and will be moving all workloads to the cloud with the blessing of the SEC. This may convince apprehensive execs to say, “If AWS is good enough for the OCC, it’s good enough for me!”
He went on to spotlight the many AWS security services, such as Security Hub, Inspector, Macie, Shield, and GuardDuty. Container runtime threat detection for GuardDuty now detects threats inside containers by monitoring OS-level activity in the container. The AWS Marketplace includes thousands (yes, thousands) of AWS and third-party security solutions.
The preview of Amazon Security Lake was a fan fave based on the cheers measured by my noise meter. I have heard it described as “the best product launched today,” but I bet the person building ETL pipelines would disagree. Security Lake allows you to combine and analyze security data to generate insights at petabyte scale. Consolidate the data from the AWS and 3rd party tools such as Crowdstrike, Barracuda, and Okta (to name a few) thanks to OCSF. Run a query using Athena against VPC flow logs, DNS queries, and firewall logs from Cisco. Managing security across accounts and systems has been challenging; let’s see how Security Lake can change this game.
High Performance Computing (HPC) at Lower Cost
The Formula 1 (F1) launch has been all the rage in Las Vegas, coming to the Las Vegas strip in 2023. F1’s racing car design simulations require over 550 million simulation points to measure the aerodynamics of their cars. They were able to reduce simulation run time by 70% by leveraging AWS.
ML inferences need high performance and can become costly. The Inf2 Instances for EC2 have 4x higher throughput with 1/10 the latency. Before choosing your instance type, make sure you do your research, as there are currently hundreds of AWS instance types available. Data and memory-intensive workloads (think wind turbine modeling) can take days to run cost-effectively, and the Hpc6id was announced as the best price-performance option for HPC.
Image above from here.
Who Doesn’t Love Massive Spatial Simulations?
AWS SimSpace Weaver can run massive spatial simulations and integrates with 3D engines. These simulations are not like classic weather forecasting but how people and objects behave in a multi-dimensional environment. Not to be confused with the spatial challenges my husband tells me I have when I’m not loading the dishwasher with max efficiency, these complex simulations could be used to model a city’s traffic flow to design an emergency disaster response plan. Developers can focus on the simulation rather than the hardware limitations of their infrastructure.
Focus on Contact Centers
Amazon Connect uses ML to do forecasting, capacity planning, and scheduling for challenges like optimizing contact center agent schedules with the right agents at the right time. Amazon Connect Contact Lens can now analyze agent performance and provide a new UI to guide agents through interactions.
Now Everyone Cares About Supply Chain Issues
Supply Chain issues have meant custom integrations and arduous processes. Amazon has a proven track record here – with their supply chain technology and capabilities of AWS infrastructure and ML, they have launched AWS Supply Chain software. The demo showcased a unified view of a supply chain that would have been helpful for my Process Management homework in B-School.
AWS Clean Rooms was announced and will fast-track the weeks it could take a company to build a new clean room, allowing collaboration with shared datasets and protecting underlying raw data. For example, rewards programs and ad-click behavior can generate insights to serve better ads while protecting the customer’s privacy.
Saving the Best for Last and Saving Lives with Data
A sobering biology lesson included the advancements in cancer treatments and how companies like Lyell perform T-cell therapy research and partner with AWS for real-time monitoring of data.
Amazon Omics (think the study of DNA or RNA) enables collaborative research at scale with privacy and integrates with Amazon HealthLake or SageMaker. This data can be captured to evaluate drug effectiveness as one use case example.
AWS listens to customer requirements, and this continues to drive innovation. What’s driving your organization’s innovation?
The AWS cloud and their services portfolio offer a comprehensive suite of capabilities to help you accelerate innovation, improve efficiency and speed to deployment, and unlock new revenue streams by utilizing, data analytics and machine learning capabilities. If you’re interested in leveraging AWS or any of their service offerings, utilizing AWS best practices, get in touch with our experts to explore how we can help you meet and exceed your goals.