After walking 43 miles in the 5-day event, our third participation in the biggest event of cloud computing finally came to an end. We were at the AWS re:Invent, one of the world’s biggest events in technology and certainly a unique networking opportunity.
Amazon Braket brought “quantum as a service” to AWS’s customers, but it is still as a preview. Users will be able to run simulations of quantum computers with partners, such as D-Wave, IonQ, and Rigetti.
There were many significant announcements about machine learning initiatives, but the highlight was Sagemaker.
With “SageMaker Operators for Kubernetes”, you can train and deploy AI models using Kubernetes. AWS covers a gap between applications in containers and SageMaker workflows, while still delivering preconfigured and optimized computing power without the need to customize code.
Among the many innovations of Sagemaker, it is worth mentioning the SageMaker Studio. It is certainly the first fully integrated development for machine learning.
There are also SageMaker Experiments that will allow developers to visualize and compare iterations of machine learning models, training parameters and results to make an easier and more effective analysis.
The SageMaker Autopilot generates machine learning models to receive data in CSV format. In addition, the SageMaker Debugger monitors real-time ML models.
Still in preview, the Amazon CodeGuru ML is a new service that automates code reviews and generates application performance recommendations.
CodeGuru detects problems, such as resource leaks, potential competition problems, among other things. Additionally, it is trained in the best practices and lessons learned with millions of code revisions in open source projects and Amazon’s own products.
With Contact Lens, AWS brings machine learning for call centers. In fact, it helps to monitor conversations with customers through sentiment analysis, trends, and other factors to train attendants and review processes faster.
The Contact Lens uses natural language processing and speech-to-text technology to analyze transcribed calls. In addition, it creates a fully searchable archive to scan transcripts for trends and insights.
They announced the world’s first musical keyboard with machine learning. DeepComposer, a 32-key keyboard, enables data scientists to test AI models.
Check out the DeepComposer in action with Beethoven!
Currently, Google has a 92% share of the global search market. Maybe that will change someday with Amazon Kendra. It is a search service based on machine learning, but its focus is companies.
Kendra responds to queries from natural language and research on different file systems, such as SharePoint, intranet sites, and file sharing services to retrieve accurate and relevant data.
By using machine learning, Kendra learns over time by using algorithms to determine which results in more value are encountered by users.
AQUA is a new and distributed cache (HW accelerated) that maximizes Redshift consultations up to 10x. In other words, the cache will allow it to work up to ten times faster than any data warehouse in the cloud.
AQUA comes to cover a gap between processing and storage by performing data processing on-site in its new cache. Furthermore, it is backward compatible with its codes in the current version of Redshift.
The S3 Access Points is a new feature of S3 that facilitates access to data at scale.
Presently, you can easily create hundreds of Access Points to S3 buckets, each with a specific and custom permissions name. With the Access Points, you can facilitate the process of creating and maintaining access to shared S3.
Access Points are available in all AWS regions without additional cost.
The CEO of AWS, Andy Jassy, said that the company would be evaluating a new service to help signal any fraudulent online activity.
Amazon Fraud Detector is a fully managed machine learning tool with decades of training in best practices and fraud detection data to automatically recognize illicit activities.
Besides Fraud Detector, Andy Jassy presented the Amazon Detective preview version. The tool is designed to simplify the analysis, study, and identification of security problems and suspicious activity.
Therefore, Amazon Detective can autonomously collect data and use machine learning to create a data set to allow users to conduct faster research.
Additionally, the service can analyze trillions of events from various data sources across the cloud to create a centralized, interactive view platform, users and interactions.
Cisco and AWS are teaming up on a mission to improve the performance and security of applications. Therefore, the alliance will give deeper integrations between Cisco network services and AWS. As a result, it will equalize the most flexibility, security, and efficiency for their hybrid networks.
One of the biggest announcements in Dr. Werner Vogels’ (CTO of AWS) keynote was the Amazon Builders Library. It is a new portal that allows behind the scenes access to view how the Amazon and AWS develops, launches and runs its software.
So, the articles are produced by Amazon’s technical leaders to give us a close look at the inner workings of Amazon and AWS.
In the end, everyone was already anxious about AWS re:Invest 2020 again in Vegas! After all, technology is at the heart of what we do at Zoox Smart Data.