The company I work for, Perion, chose Amazon Web Services as a main cloud provider for managing and operating our applications. These days I am learning how to manage databases and data related flows in the AWS cloud.
Which is the reason I have attended the AWS summit in Tel Aviv today.
Today’s AWS summit was the first out of the series of AWS summits held in major cities around the world. There were ~ 5000 attendees today, 25 technical sessions and it was sponsored by 33 AWS partners. Most sessions that I have attended today had a lot of sponsor content mixed with technical not-too-deep dive into various AWS services.
Sessions were divided into 5 tracks, led by AWS Solution Architects and Sponsors representatives:
Big data & Analytics sessions on data ingestion, data transformation and data analysis using AWS services
AI and Deep Learning – sessions on AWS offerings for data enrichment and learning models
Innovate – Rapid Application Development services
Cloud operations – sessions on DevOps methodologies and tools
Let’s start - for those who starts their journey into the AWS cloud
I have really enjoyed the keynote by Dr. Werner Vogels, Amazon CTO. The keynote hall was already fully packed when I have arrived, and I had to watch the keynote video stream from the second hall. The sound was awesome but the person who was operating the camera wasn’t well at his job. Half of the slides could not be seen; the focus was all the time on the speaker.
Dr. Werner have talked about Amazon revenue run rate, which is near $20 billion dollars ( this forbes.com article says $16 billion, Microsoft has $16.7 billion, IBM $15.8 billion). AWS have millions of active customers and their pace of innovations (amount of the unique services per year) grows at very high pace:
24 during 2008
48 during 2009
61 during 2010
82 during 2011
159 during 2012
280 during 2013
516 during 2014
722 during 2015
1017 during 2016
1403 during 2017
Amazon for 7 consecutive years named as a Leader in Gartner’s Infrastructure as a Service (IaaS) Magic Quadrant.
Dr. Werner have talked about how infrastructure and software vendors used to tell us how to develop software. These days, in the cloud, we have as many tools as we need to make our development tailored to our needs.
Modern applications are continuously evolving. They have more than web content, they are using micro services architecture and are api-based. Everything is programmable, many have a smart connectivity, push notifications and responsive design. We used to have 1 deployment a week, now we have 50 deployments a day.
There were many awesome examples of the AWS customers during the keynote. One of them was the cloud infrastructure behind the iRobot:
(Image source : https://aws.amazon.com/solutions/case-studies/irobot/)
Dr. Werner have mentioned how Netflix are making use of Machine Learning to build smarter applications. 75% of the Netflix videos were viewed after the recommendation made by the recommendation engine. He has also mentioned the Pinterest, one of the world’s largest visual bookmarking tools, where all images go through Machine Learning process to get a better grip on visual search. Another cool examples were the boxing gloves that check the punch speed and the punch intensity, the helmet that analyses the pilot’s emotions and soccer and baseball video ingestion that uses Machine Learning to help coaches to understand how the team can do better at the next game.
Machine Learning can be used in many more use-cases:
Shopping advises
Customer problem detection
Product modeling
Production substitutes
Customer adoption models
Sales lead ranking
Dr. Werner have talked about the modern data architectures and their 10 main characteristics: reliable data ingestion, preservation of source data, lifecycle management, metadata capture, security, self-service discovery, data quality, analytics, orchestration and capturing the data changes. He has mentioned that 60000+ databases already have been migrated into the AWS cloud from private datacenters. He has also talked about Aurora, the fast-growing database service. The Aurora data is being copied into 3 regions and has up and down auto-scaling capacity. For example, Expedia that has 4000 writes / 5000 reads per second into the Aurora database.
I will skip the keynote part where Dr. Werner have talked about rapid application development, serverless architecture, code services, kubernetes and application services – that is not my cup of tea.
There was an awesome demo of Trax, the company that uses innovative image recognition technology to help retail stores to sort out their shelves and are using Artificial Intelligence to unlock business opportunities. As a next stage of their solution, they are planning to use robots to take pictures inside stores which is very cool.
Another great demo was about Flytrax, first drone based food delivery service. Drones are super-fast, the can fly up to 60 km/h, affordable, not impacted by the traffic and can easily reach almost any location within a few minutes instead of driving for a few hours. Unfortunately, this service is not available in Israel yet.
Great keynote, #awssummit !
Yours,
Maria
