Re:Invent 2019 – Exciting News For Qrvey Customers

AWS surprised us again this year with literally hundreds of new features and services across their platform. And as usual, we’ve been combing through every announcement to pick out which ones are the most relevant for the next generation of analytics applications and which ones will be integrated into the Qrvey platform. We previously mentioned some of the machine learning advancements, but here are a few other nuggets that caught our attention.

First up, we were delighted to see ElasticSearch getting some love. AWS has always been obsessed with efficiency and reducing costs, and they know that storing data at scale can be expensive. The only alternative is to store less data, but that leaves your metrics and analytics incomplete and inaccurate. Fortunately, now there’s a new alternative called UltraWarm which promises 90% savings over the hot storage options we currently use at Qrvey. Since Qrvey deploys straight into your existing AWS accounts, these savings would flow directly to your bottom line. We won’t get into all of the technical details here, but UltraWarm looks like a promising new data storage alternative that we’re already exploring.

Many of the other AWS announcements we found most intriguing this year surrounded machine learning. Altogether, Amazon made 20 top-level machine learning announcements, showing their commitment to making this powerful new technology more accessible to developers across the platform and ultimately to end-users. This is great news for Qrvey because it will allow us to bring more ML features to our platform in an even faster way than initially  planned. That’s the power of being a cloud-native analytics platform. Since we’re already built on microservices, we leverage AWS like no other, adding and upgrading individual components faster than anyone.

What are some of the exciting things users will soon be able to do with these new ML-enabled services? Amazon showed off a new fraud-detection service powered by ML as one of the examples. After submitting a simple training set of data, the service will get to work, providing all future transactions with a fraud score that can be used as part of your analysis and decision making, all without having to touch the ML layer directly.

We can’t wait to dive into all of these great new announcements and see what possibilities for analytics lie in 2020 and beyond.