Kinetica in Motion
The Kinetica Blog
Blog »
Amit Vij

Embracing Real-time Streaming Data

Tweet about this on TwitterShare on LinkedIn0Share on Facebook0Share on Reddit0Share on Google+0

The data connectivity world is changing rapidly, with a firehose of data streaming in from devices, people, and channels. How can organizations act on all of that data? We created Kinetica for those competitive organizations who need to leverage big, fast, data in order to remain leaders in their industry.

How did Kinetica come about in the first place? Let’s take a few steps back so I can explain what’s happened to data management infrastructures over the last several years. In the past, organizations have struggled to benefit from the influx of live data such as IoT, as their legacy systems were only designed to periodically load data from transactional systems (OLTPs). They were hitting the performance and scale limits of today’s in-memory and traditional databases; they simply couldn’t ingest and query multiple real-time data sources.

We created Kinetica, a GPU-accelerated database that simultaneously ingests, explores, analyzes, and visualizes streaming data, to solve the real-time data problem.

I started Kinetica (formerly GPUdb), along with CTO and co-founder Nima Negahban, from the ground up. In 2009, the United States Army Intelligence and Security Command (INSCOM) wanted to be able to track national security threats in real time. They needed a solution that would ingest and fuse multiple high velocity data feeds, filter and perform advanced analytics, produce real-time results, and visualize insights globally.

We started by evaluating best-of-breed technologies across the world; we learned the pros and cons of each and their pain points. INSCOM had acres of Hadoop nodes. After their relational databases failed, we used this opportunity to test every flavor of NoSQL database out there. Eventually, our military leadership even created their own NoSQL database, Apache Accumulo. However, none of the existing technologies worked sufficiently; data warehouses were too slow, NoSQL wouldn’t scale, and in-memory solutions promised real time but didn’t deliver. Clearly, there was a major technology gap, and this was the very reason INSCOM incubated us.

Starting as newbies on a major military project where other teams were a lot larger and more connected, we grew through the ranks over the years to have the privilege of working for top US military leaders, and we were able to achieve game-changing results.

The database that we developed was centered around massive parallelization, utilizing GPUs to explore and visualize data in space and time–an approach that has been patented and won a High Performance Computing Innovation Excellence Award from IDC. And then in 2012, GPUdb was launched at scale, ingesting and analyzing over 200 different data feeds, to track terrorist activity in real time.

In 2014, we launched GPUdb in the commercial space. We realized that businesses were running up against the challenge of streaming and analyzing data in real time, and we knew Kinetica could solve this problem. Our first commercial customer, the United States Postal Service, deployed GPUdb into production to optimize routes and increase accuracy. In fact, we won our second High Performance Computing Innovation Excellence Award from IDC for our innovation on behalf of the United States Postal Service.


Fast forward to today: we have re-branded the company to Kinetica and moved our headquarters to San Francisco, so that we can focus on being the leading GPU-accelerated database provider for those organizations that need to make critical decisions instantaneously based on the influx of their data.

To add to this momentum, we recently added $6 million in acceleration financing from Ray Lane, former President and COO of Oracle, for a total of $13 million, which is fueling our entry into the commercial space after 6 years of development and deployments for the government sector. We have also expanded our team across the board in operations, sales, marketing, engineering, and services, bringing in top talent from big data leaders such as Cloudera, Hortonworks, MapR, and Platfora to drive our growth. Our recently announced Version 6.0 of Kinetica features advances in performance, visualization and high availability for enterprise-grade stability and security.

Keep an eye on this blog to learn more about trends in the big data industry, how to incorporate streaming and IoT data into your organization and ways other organizations are using Kinetica to conquer supply chain, logistics, fraud detection, regulatory compliance, customer 360 and a wide range of other business challenges.

Leave a Comment