>> Hey welcome back everybody. Jeff Frick, with the Cube. We're at Node Summit 2017 in downtown San Francisco. About 800 people, developers talking about Node and Node GS. And really the crazy adoption of Node as a development platform. Enterprise adoption. Everything's up and to the right. Some crazy good stories. And we're excited to have somebody coming right off his keynote. It's James Bellenger. He is an engineer at Twitter. James, welcome. >> Thank you, thank you for having me. >> Yeah, absolutely. So you just got off stage and you were talking all about Twitter Lite. What is Twitter Lite? I like Twitter as it is. >> Ah, so Twitter Lite is an optimized, it's a mobile web app. So if you pull up your phone, open up the web browser and go to twitter.com, in your smart phone web browser, you get a Twitter experience that we're calling Twitter Lite. >> Okay. >> And it used to be a little bit out of date. But we've been able to update it using a lot of new exciting web technologies. And so now we have this thing that feels very much like a native web app. >> Okay. >> They call them progressive web apps these days. And so we're using that as sort of a way to sort of compete in areas and markets where maybe a native apps are less able to compete. Where you know, people don't want to download a 200 megabyte iOS app. They want something that fits under 600 kilobytes. >> Okay. So you had the Twitter Lite app before. And then this was really a re-deployment? Or am I getting it wrong? >> I think, well we had We had a web app at mobile.twitter.com. >> Okay. >> And it was just sort of the mobile web app. >> Okay. >> But you know we sort of really rewrote everything. And that includes the back end on Node. And then we're now sort of pushing that and calling it Twitter Lite. >> Okay. And when did that go live or GA? >> About three months ago. >> Three months ago, okay. Super. So obviously you're here at Node. You just spoke at Node. You know, how was the experience using a Node tool set versus whatever you had it built on before? >> It's definitely faster in every way. Well, I mean, >> Faster in every way. That's a good thing. >> So well, let me Let me catch that. Be more specific. It is ... >> It's those benchmarking people. We need them back over here. >> It is very fast for how we apply it. It's really fast for development speed. And perhaps the biggest win is that on both sort of areas of our stack whether it's the part of the application that runs on the browser or it's the part of the application that runs inside the Twitter data center. We have one language and technology. So when a problem comes up and an engineer needs to like go and find the problem and fix it they don't need to sort of "Oh, well that's server code. "I don't know how it works. "And it's written in this language I don't understand." We really just have one application and it happens to run in both places. And so it really improves engineering efficiency. >> And you saw that in the development process, QA and the ongoing. >> Yeah. >> Yeah. And was it more ... So it's more like the guys that were more front end that now have access to the back end and then the other way around. Is that correct? Yeah, it's a little bit of both. >> Okay. >> You know, I think before I think there's people that they really like Scala. And they only want to work in Scala. Or there's people that really don't like it. So you end up, I think, having engineers kind of get bulkanized by their technology choices, and their preferred systems. But I think it really sort of tears down a couple walls. And so it makes, it improves engineering efficiency that way. But we found also that some of the tool sets and the tool chains that we're using allow engineers to just sort of like move faster. >> Right. >> So you know, whether that's like recompiling the service in like one second. Instead of having to wait for multiple minutes. There's just sort of less time spent waiting. >> Right. And in terms of don't share anything you're not supposed to share but in terms of, you know, frequency of releases and ongoing maintenance and kind of the development of the I won't say the app, not the app. I guess it is the app. Going forward, you know, how has that been impacted by moving to this platform? >> I think it might be too early to say. >> Okay. >> We've, you know, right now we've got about 12 to 15 engineers and we're ramping up. And it, I think it might, we're kind of looking to finish around 25 engineers, by the end of the year. >> Okay. >> So the sort of team and contributor base of the kind of like core team that are working on the app is growing. But you know, otherwise there's, you know, we're releasing every day. We're, you know, we try to you know, we're always pushing code. We're running experiments a lot. >> Right. I don't know if that answers your question but. >> So it sounds like it's a little easier but you're still doing everything you were doing before but now it just feels like it's easier because of this. >> Well, you know, talk to me in a couple months. >> Okay. >> Then maybe we'll have some better answers for you. >> Okay. So the other thing I want, if I talk to you in a couple months, I talk to you a year from now, just in terms of as you look down the road, you know, what this opens up. You know, kind of what are some of your priorities now that you've got it out. You said you've been out there for three months. What's kind of next on your roadmap, your horizon? >> So far, I think we've been really encouraged by the success of using this stack for development. So we're looking to kind of double down on that. >> Okay. >> So that means looking at some of the other Twitter web apps. Oh, sorry, Twitter apps in general. The other ways people use Twitter. And to sort of look at how they were built. And to see, because we're using React, and because we're using, I think technologies that make it very easy to you know, be responsive and you know, either be have a wide layout or a very narrow layout, or work offline. We have a lot of potential to sort of cannibalize or replace and also update some of the existing apps >> Right. >> That maybe don't get the attention that they need. >> Right. >> So there's some of that. And then I think Twitter Lite as a product I think that we're going, you know, we're looking to really expand it's reach. And make a big push in some of the developing areas. >> Yeah. Because the other thing people don't know, I mean, Twitter's acquired a bunch of companies, you know, over the years. So we've heard some examples earlier today, where that's a use case when you do have the opportunity to maybe redo an acquired application. You know, that those are kind of natural opportunities to look to redo them with this method. >> Yeah. Sure. >> All right. Cool. Well, James, thanks for taking a few minutes. >> Thank you. >> Congratulations on the talk. And I'll think of you next time I go to Twitter Lite. >> Yeah. Thank you so much. >> All righty. He's James Bellenger from Twitter. I'm Jeff Frick. You're watching the Cube from Node Summit 2017. Thanks for watching. (techno music)