What has AWS been doing in 2020?

In this article, I'm going to summarize the major announcements and product release during the 2020 AWS re:Invent.

Quicksigth Q

Quicksight is an AWS alternative to Tableau as a Business Intelligence tool. The "Q" is an additional feature that uses Natural Language Processing (NLP) where user asks questions and query is built in automatically. Next time ask "What are my most ordered products?" and you will be surprised how your question gets translated into SQL based query and the result is returned automatically.

EC2 now supports MacOS

Finally MacOS 10.14 and 10.15 can be run on EC2 without using the traditional virtual machine approach!

Babelfish for Aurora Postgres

Surgical strike to Microsoft, helps in easily switching from MS SQL Server as it can understand TSQL and SQL and migrate over to Aurora Postgres.

Aurora Serverless with no cold start

With Aurora Serverless the users did not have to keep a database instance running 24/7 but be used based on demand. But earlier there was a problem when the first call to this database would take between 5~50 seconds but now it can perform instantaneously.

AWS Proton

This aims to streamline the deployment process for serverless and container based microservices.

AWS Lambda memory increase from 2gb to 10gb, billing in milliseconds and support containerized lambdas

Huge changes to Lambda services as memory can be increased for performing complex jobs and should be a good use case for Machine Learning. Earlier the billing was done in 100 milliseconds interval but now customers are billed based on the milliseconds used which can result upto 70% cost reduction. Finally there is Container based Lambda based environments support.

AWS Glue Elastic Views

This is one of the most exciting feature, earlier if a developer wanted to get their database from DynamoDB into Redshift for analysis, they would have to build a datapipeline to listen to DynamoDB streams, input that data to firehose, deliver that data to s3, and then finally initiate a Redshift S3 Copy command to move that data into Redshift. This was a nightmare.

With AWS Glue Elastic Views, AWS is now taking care of this complexity for free. Now developers can ‘point and click’ their source and target databases and watch the magic migration happen. Possible sources include Aurora, RDS, and Dynamo, and Targets include Redshift, ElasticSearch, S3, DynamoDB, Aurora, and RDS.

AWS Sagemaker Data Wrangler, Feature Store and Pipelines

Sagemarker Data Wrangler solves the data preparation problem that comes with machine learning. Feature store acts as a repository for developers to create, store, and share features to be shared across multiple different ML based applications. Sagemaker Pipeline is the first exploration of CI/CD pipelines for Machine Learning.

AWS DevOps Guru

With DevOps Guru, AWS is taking it one step further of AWS Code Guru, looking to identify operational problems before they impact customers. Not only do they recognize these problems, but they actually suggest recommended actions to fix the problem

AWS Connect

With the Covid 19 pandemic, the need for call center agents to work from home is larger than ever and with AWS Connect make Call Centers even easier to build, manage, and operate. Some major announcements to this sphere:

• Amazon Connect Wisdom: This feature aggregates data from your data store for your agents on the fly as the call is happening. Agents no longer need to fumble with multiple tools and databases to acquire relevant customer information and can instead have that information presented to them automatically.
• Amazon Connect Customer Profiles: With this feature more context into your customers can be proactively acquired before and while a call is taking place. This means pulling up a history of past calls and aggregating all that information to your agents.
• Amazon Connect Task: This feature lets managers create, assign, and track tasks for their agents. They can even assign tasks based on how busy their agents are, or how much time left they have in their shift. 
• Amazon Connect Voice Id: This mind just be the most mind-blowing and innovative feature. Don’t you hate calling in somewhere and reciting your name, address, and everything else? I sure do. Connect voice ID looks to build a audio footprint of your voice so that you can be automatically identified by the agent. No more annoying questions when calling in!

AWS Lookout

Amazon lookout provides anomaly detection, sound detection, vibration assessment, and temperature detection for predictive maintenance of hardware.

AWS ECS and EKS Anywhere

With ECS and EKS anywhere, AWS is bridging the gap between cloud based and on premise workloads. The idea of this service is to allow you to run the ECS and EKS frameworks using your own metal and in your own datacenter.

New Gravitron c6gn instances

Gravitron is AWS take on building their own CPUs. It can deliver 100 Gbps network bandwidth, 38 Gbps EBS and has 4x performance in many customer use cases.