 Cube on the Ground at Spark Summit 2015, and I'm here with Manish Amdi from Origami Logic. He's a software engineer, and I actually found him through the crowd chat back in platform searching the Spark Summit, people that were on Twitter and that were influencers, so I reached out, brought him on the Cube, and wanted to find out how he's implementing Spark. So, can you tell us a little bit about what you do at Origami and how you implement Spark? Thanks for having me here. Yes, I work at Origami Logic. We are a cross-channel marketing intelligence platform where we cater to marketing organizations that have thousands of digital marketing channels across multiple mediums such as social, web, CRM, what have you. What we do for them is we collect the data automatically on their behalf. We help them visualize the data in a language that marketers understand, and we also do analytics for them. It could be simple analytics such as doing custom APIs, to very sophisticated calculations such as doing aggregations over millions of marketing objects, or doing distributed data science where we're going to find out anomalies or dimensional insights for over thousands of time series. So my role at Origami is to actually implement the Spark-based backend, which are less to a tremendous improvement in speed and performance, and help us bring real-time information into our system. Awesome. Sounds like something that CrowdChat would be interested for all about the social atmosphere, keeping data within the social realm and seeing what's going on. So how do you personally use Spark and what do you think the contributions to Spark are having on the technology sector in general? So I've been a Spark ML Lib contributor since two years ago. I've personally contributed some core ML Lib features, which is the machine learning library in Spark such as decision trees, random forest, gradient boosting. So given that I'm partial to Spark and I've seen its benefits right from the start, at Origami we use Spark for three projects. One is a real-time distributed query engine, which can answer queries very fast in a low-latency, high-throughput fashion. We're also using it for our social listening needs, where we have a lot of data coming in from Twitter, Facebook, and other social platforms into our system, and we are creating our own custom and analytics on top of it. And thirdly, we have a new feature called Origami Stories, where we are actually finding out insights on behalf of the customers using machine learning algorithms and bringing it out into our product both as a marketing story feed, such as a Facebook story feed, but for marketers, as well as a notification system to notify them when something is going great or wrong in their own digital space. Interesting. It sounds like great real-time use cases. Awesome. So what do you think of the conference here today? It's been great. I spoke here last year and it's grown quite a bit. It's good to see new faces as well as the old ones still contributing as they used to, looking excited towards the development of Spark over the next couple of years. Yeah, I'm sure it's a small community. You're building it as you go. So last question, I found you through social media. What are the benefits of being active on social? I know that Origami is active on social, and you guys look at the data. Can you talk a little bit about how that might benefit business and the rate of interest on return, the rate of return on your investment and time? I see. I use social media in an organic way. I just follow people who are working in the big data space personally and people find me accordingly. We have a lot of impromptu discussions going on. People have applied to Origami Logic by finding us through Twitter and Facebook. So things have been going great. It's more of an organic way of communication, but it's turning out much better than we expected. All right. Well, that's a wrap for today. And we're here at Spark Summit 2015, Cube on the Ground. See you later.