Our team attended the AWS re:Invent conference in Las Vegas from Dec 2 to Dec 5. More than 65K people came to the conference (vs 50K last year) and there were more than 3K sessions (vs 2.1K last year). During the four-day event, we attended all the four keynotes by Peter DeSantis (VP of AWS Global Infrastructure), Andy Jassy (CEO of AWS), Doug Yeum (head of AWS Worldwide Channels), and Dr. Werner Vogels (CTO of Amazon.com).
In the main keynote by AWS CEO Andy Jassy, the company announced 23 new features to address customer demand across all technology stacks, including hybrid cloud, ASIC, data lake, AI, and serverless computing. Andy Jassy emphasized AWS’s lead over other public cloud competitors, especially over Microsoft. On one occasion, he even criticized Microsoft’s licensing practice, which limits the ways businesses can deploy Windows and SQL Server software in the cloud using existing licenses.
In the keynote by Werner Vogels, he detailed the behind-the-scenes hardware and software innovations that make AWS more performant and secure, such as AWS’s virtualization journey powered by “AWS Nitro System” and “AWS Firecracker”. Dr. Vogels also emphasized that enterprises are rapidly adopting serverless. There was one session that showed how Vanguard uses AWS Fargate (serverless compute engine) to lower its cloud spending by 30%.
In addition to the keynotes, we visited some technical sessions and expos that interested developers the most. Below we summarize our takeaways in two parts: the first part is a snapshot of the most important new feature announcements from AWS re:Invent; the second part is our takeaway from the technical sessions.
Part 1: new feature announcements
Part 2: tech-sessions takeaways
5G Related
Amazon announced that AWS will partner with Verizon to improve 5G speed on December 3. The service is called WaveLength, which uses 5G networks to transfer data from the cloud at a faster speed. The idea is similar to many other edge computing technologies: it places small data centers running AWS software next to Verizon’s 5G points of presence. The results will be ultra-low latency for AWS applications.
In September’s Alicloud’s Apsara conference, Alicloud formed a similar partnership with China’s telecom companies. Also on November 26, Microsoft announced that AT&T integrating 5G with Azure to enable next-generation solutions on the edge.
Amazon Connect
Cloud contact center software is one of the fastest-growing businesses. Companies such as Twilio and Zendesk are great successful examples. In Dreamforce 2019, AWS made big news when Salesforce announced that its “Service Cloud Voice” will use Amazon Connect to provide AI-based contact center solutions. During re:Invent, Zendesk also announced a strategic, go-to-market solution package based upon an enhanced integration between Amazon Connect and Zendesk Talk Partner Edition.
In addition, AWS announced “Contact Lens for Amazon Connect”, a new machine learning-powered analytics for customers to better assess call contacts. We attended one session that featured AWS, Salesforce and John Hancock. The demo is very similar to the state farm demo during Dreamforce 2019.
Datawarehouse: Redshift versus Snowflake
Analytics tools of Datawarehouse (structured data) and Datalake (unstructured data) are big battlegrounds for both cloud providers and emerging software companies. For example, in Microsoft Ignite 2019, the company announced Azure Synapse that brings “limitless analytics service to enterprises for both data warehousing and big data analytics”
For years, Snowflake is taking market share in the cloud-native Datawarehouse market even it is using AWS as host. At re:Invent, AWS announced a new analytic tool, AQUA Redshift, which claimed to have 10x the performance versus competitors. We will follow up with its impact on Snowflake.
One session featured Nasdaq’s experience with Redshift. Nasdaq IT department has deployed Redshift since 2013. But as data volume exploded (50 billion transactions per day), Nasdaq had problems with Redshift’s capacity (up to 128 notes) and costs. The Nasdaq IT team contacted the AWS data lab. AWS put together a team with its top data scientists and came up with a solution for Nasdaq within four days, solving the problem that took Nasdaq many years without any success.
Netflix: open-sourced Metaflow Python library. On December 4 at re:Invent, Netflix's data science team open-sourced its Metaflow Python library. Metaflow was built internally to help boost the productivity of Netflix’s data scientists who like to express business logic through Python code but don't want to spend too much time thinking about engineering issues. The announcement drew many attendees and the Venetian Theater was occupied by developers.
Zoom and Splunk
Zoom uses its own data centers (colo) for its real-time conference features. However, it uses AWS for its cloud IT, such as log-ins and identity management.
Zoom started with a 20GB Splunk license and currently scaled to 500+GB.
The Splunk's volume-based indexing fee is discounted significantly as Zoom only uses part of Splunk’s offerings.
Esoterica's statements are not an endorsement of any company or a recommendation to buy, sell or hold any security. For full disclosures, click here.