Lightbend
is a mature startup based in San Francisco that is about to celebrate its 8th
birthday. Originally known as the home of the programming language Scala and
middleware software Akka, today Lightbend offers two platforms for the
enterprise: the Reactive Platform and the Fast Data Platform. With more than
150 enterprise customers, and Web giants like Airbnb, LinkedIn, Lyft, Shopify,
Spotify and Twitter running open source Lightbend software in their core
infrastructures, the company knows about scale and fast data. Going into the
big data conference Strata, they are announcing the latest release of the Fast
Data Platform. I recently spoke to Lightbend CEO Mark Brewer to learn more.
VMblog: To kick things off, can you talk about and explain the Fast Data Platform?
Mark Brewer: Our Fast Data Platform basically wraps the
Reactive Platform with additional technologies to help enterprises take data in
motion in streams and marry it to their business applications, typically tied
to microservices. We bring other streaming technologies to our platform like
Apache Spark, Apache Kafka, Apache Flink as well as our own streaming
capabilities. We pair these with the rest of the Lightbend stack and offer it
as a single, commercially supported distribution for building these new
applications that rely on streaming data. We marry microservices to streaming
data.
VMblog: Can you tell readers what's new in Fast Data Platform 2.0?
Brewer: Probably the biggest news is our support
for Kubernetes. Kubernetes
and Fast Data Platform is the perfect marriage between distributed computing's
most production-grade container orchestration solution in Kubernetes, and the
most streaming data-optimized programming abstractions and run-time tooling in
Fast Data Platform. In addition to native integrations with Apache Spark,
Apache Kafka, Apache Flink, Akka Streams and streaming data's other hottest
frameworks -- Fast Data Platform provides the management tools and production
monitoring capabilities that allows operations teams to run and troubleshoot
streaming data systems in production.
VMblog: And what can you tell us about use cases and customers already on the Fast Data
Platform?
Brewer: I can tell you about four customers very
familiar to your readers. First is HPE, what used to be Hewlett-Packard. They
have a massive IoT system. They wanted to give their customers for storage
devices and other hardware all the metrics and analytics they need to know how
the equipment is performing and if something like a hard drive is about to
fail. In this case we're talking about literally 20 billion sensors generating
seven million messages or metrics every second and that's being processed over
our platform.
In the cruise industry, personalization is
critical to keeping customers happy and retaining their business. We have a
number of customers here, including Royal Caribbean, who use our platform to do
everything from making a reservation at a restaurant to booking a massage or an
outing like snorkeling. There are unique
challenges here. Our system has to work even when it's not connected to the
Internet!
Another interesting use case is CapitalOne,
the big bank. We are the platform that runs their app to help car buyers get
financing and find the car they want to buy or lease. This system links more
than 12,000 dealers across North America and is used by more than 4 million
customers. Buyers can get finance decisions literally in seconds to see if they
qualify for a loan to buy the car they like. It's orders of magnitude faster
than their old system.
Finally, Starbucks rolled out its loyalty system and
coffee ordering app on our platform first in Japan and now it's in the United
States. Additionally, Starbucks has rolled out it's new point-of-sale system
built on the Lightbend platform.
##