 Yeah. No, you're not. You're fine. You can stop all at the launch. Thank you. We're live. Thank you. We're live. We're live. Thank you. Hey, hey, hey, hey, hey. What are you doing? All right. Okay. You're welcome. Can you hear me? No, no. All right. No, I'm all right. Thank you. All right, I think just say no. Thank you. And we're going to introduce us to you. What's your name? Oh, yes. Ladies and gentlemen, welcome to Rolbs. I can do this. I can do that. I can do this. Yeah. Really, really, really good. Ladies and gentlemen, put your hands together. Welcome. It's awesome. Here's your CEO. Here we go. See you. See you. Happy. Happy. You did a great way to base it. I'm on the wall. I'm on the wall. Please, please, please. People will love it. It will be hilarious. This is great. You did the intro. Yeah. You did it. You're good. Reach. You can do the heel. I just want to push this off. Man, tell me when you feel this thing. One, two. So, I'm trying to get this one on my arm. Tell me when you feel this thing. One, two. One, two, three. One, two, three. You're OK? You're good. All right, for the rest of the day, I have a couple more things to say here. And this one is for you. Well, we're ready to go. We're going to do this thing now when I sit here. All of you, you can do it. We're not seeing things, but we're not seeing things. Hey, Phil, I think we became strings, so we've got to make sure that Steve on the home base is the exo-state because if you go from the home base, you get to the right string. You guys are live, right? Yeah, but there are two different strings. One string will find you to the wrong string, and the whole string should have you to check that it's another string. It's like this. Oh, sorry. I told you, the way that I listen to music right now is to use this one. It's the wrong one. We don't have five people watching. Okay, where would this be? This part. I'm going to be sure to give you guys a big one one time. Cool. Thanks for that. Who's doing the tap, people? Sorry. I'm just going to switch the stage. You have to make that. What's your point of view? Who's doing the tap? Who is your tap? Who is it? It's you. It's not you. It's us. Okay, what are we doing? Oh, we've got to fix that. We're helping you out. Hi. And there's a 40-second video. What's that? It's from this. It's from Google. It's from Google? Yeah. Okay. Cool. Thanks. Don't you guys know that? You should have a good connection. No. I don't want to stay with you. I think you need to meet your wife. You can watch your slides here. And you have a follow-up call. Okay. Okay. Please work with your data. Okay. All right. Can we go live? Yeah. We're doing the links everywhere now. Okay. So we're waiting for that. So it's the after slide. We will give you a link to the correct screen. That's what we're waiting for. Okay. And I think you can contact us during the session. You can go on when you get to the slide. Please have a look before you come off the car. Okay. Go ahead and exit presentation mode. And then lower that hand to the presentation. No, we're not going to do that. No, we're not going to do that. Okay. And we're going to take the picture. We're going to zoom it in on it. It's like five minutes before the second. I'm not going to do that in the future. Okay. I'm not going to do that. Yes. We're going to go, Okay. We're going to take the picture. Okay. Okay. It's too nice. Yeah, we're going to go here. You're just in the picture of the. Yeah. No, it's great. Okay. The brilliant are the three letters. All right. The room is going to be uniform. It is there. One is. I want to hear anyway. So, so people will come up on time. I want to hear anyway, so the people will come outside. Okay. My phone is not free right now. Is there anything we can do about it? Do you want to? Yeah. Do you want to be a police officer? Um, Just hold it off for a minute. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. Okay. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we can hear. We have a lot of reverb and echo from people on the live stream that are asking that we We interact with each other, we see each other's chat messages, the comments, we do video calls, but we have a culture of directness, so it's not always a pleasant message that is brought to you, and it helps if you have a bit of, you get to know the person on the other side, you establish your credibility with each other, and that's where this something is for you. And when we had the first summit in Serbia, we were kind of jokingly referring to it as a good-luck summit, because it was just a few of us. But over the years that closed, and the mountain view during YC was a three-month summit, it was a bit too long to be together, we didn't even say not for that amount of time that we get a warmest. We need to go back home, people decided they were the only ones who were warm from that time on, but there is a good time, there is an ideal time. An ideal time is about a week. And the next summit in Amsterdam was great. Chad, the person that our CRO and the person who announced me said, and said it was great because we get to roam around in the city, and that was awesome. Awesome was great because we grew so much. We've been in practical respect to time, all these different things, because we have to move people in, because we weren't expecting that big of a crowd. The last summit in Cumbia, many of you have come here, it's awesome. I see it with a very scenic location. And I hope this year we'll have a great summit again. I used to be nervous, it won't be as good as the last one, but I think I'm blessed that the nervousness is behind. It's getting better every single time. So now I want to invite Dimitri on stage to tell us how it... ...GitLab has started. So please welcome the original author of GITLab, my co-founder and our CTO, Dimitri. Hi everyone. I'm going to tell you a story about 30 days of GITLab. I have a quite long, strong connection with... ...the world of GITLab already, and shared with you. Basically the story is the best way to do it. So I'll just get started. I was working as a supplier in the world of GITLab. In the world of GITLab, I had a degree, a record degree in car engineering from the years earlier. But I liked the way development was and fixing cars. So here I am working in software company. In the meantime, I was living in a small house near the countryside. It was a really old house and like some basic things like running water. So every morning I get to the market, go to the public world, get some water for cleaning or washing. But it was not my main problem like that. I was not satisfied with the available self-adjusted tools for the market. I used github.com for open source project and it worked really well. And then in a day-to-day work in the company, I used other tools that were installed in the company. I really suffered from user experience from lack of features. So instead of doing code in your mind, I actually downloaded it and did it all fine. There was one co-worker who understood my problem. His name is Valerius Zolt and he is now working in GITLab. I and Valerius usually have a coffee break during a work day where we discuss different topics around tech development. One day we had a conversation about the cost of tools. And we were wondering, is it really that hard to put a button to fight for user story management? In terms of application, we were 100% sure we were going to build it. So challenge accepted. We worked here in the unions and we did this because we had a daily job. We were rails from governments back then and since we built rails, you can put a type or you can test. So each other version was ready in a month. We decided to open source it because we wanted to ask people to also use it and modify according to their needs. I even made a set website for it and shared it with the company. It helped us really well in a way of promotion and bringing new contributors right into the company. Because it was a hobby, we didn't plan to build a company or something like that. I continued with my daily job. It was certainly fine for me. But there we followed some of his family. But because of human leaders, I saved this project for him alone. I was really surprised by the few months more than 50 people contributed to the project. A few months later, Drivebox offered me a free server for different development. I used this server to develop GitHub. And you might know it right now at www.github.org. Because I used GitHub to develop GitHub, and actually it was in partnership with Google every time. So I came to work on this project. I did release every month in 20 seconds. The date was a random date, but the month's schedule wasn't perfect. Slowly, I realized it was important to me. The only way I worked was using GitHub. Some of my friends started using GitHub. It's a company store. In the meantime, I get an email from SIT. He said that he admires GitHub, and it's quality. And he said before I learned his contributions, really fast, I did it in one date. He also said that he is going to be there. So he started a business based on GitHub. Because of his honesty, and because he actually contributed to the project, I was really inspired by his idea. I didn't ask for money or anything. I said, yeah, keep up. I also started another project back then. It was with the API. So by the end of 2012, I had two projects to maintain. But there was also a downside. There were more and more contributions, more and more users, more and more issues. It took much more of my time. And it was really hard to combine with my daily job. And he said that in order to focus on easy GitHub, on my full-time job, yeah, I need to choose. I really like to work on GitHub, but then I need to pay a sustainable income. So I tried to make money with GitHub. First, I started this business. It didn't work well. In the best months, it was less than $100. I also tried to install and update GitLab for other users. But there was a conflict there. In order to make money on installation and update process, you need to make it actually painful and not efficient. And instead, I focused on improving installation and update process in GitLab. So I totally failed in my goal to make any reasonable amount of money with GitLab. One day, I was really, really tired. And I opened Twitter. I wrote to it. I want to work on GitLab full-time. I didn't expect any outcome of it. But a few hours, maybe later, Cid contacted me. And he proposed me to pay a salary in order to work full-time on GitLab and help him with his software as a service version, gitlab.com. I was really, really happy with the offer. But there was a risk of leaving a well-paid job. So I asked him for an upfront payment. And because we didn't have any contracts, I asked him to send money with Western Union. And it really happened. So Cid sent money to some stranger from Ukraine. He met over the Internet. It's like classic fraud. So since January 2013, I started working on GitLab full-time. It was a team of three people. Me, Cid and Marin. But Cid and Marin also worked for other companies so they can earn money to pay me a salary and cover infrastructure costs. Still, we shipped much, much faster. Every next release was bigger than the previous one. By that time, there was only one problem. We never met in person. We worked fully remote. And Cid decided to fix that. There was a conference, Rails conference in Poland, and it looked like a good opportunity for us. So I bought a plane ticket. I got my visa. It was my first international flight. It was my first flight at all. So I fly to Poland, Krakow. And we agreed to meet at Airbnb Place. And my flight was in late evening. So I arrived to the city almost by midnight. I was quite surprised when I reached the place. I haven't seen Cid or Marin there. There were some strangers. And then a few minutes later, I got a call from Cid. But before I can hear anything more than hate beatery, I run out of money on my prepaid phone. So I was in the foreign country at midnight in some living district without a smartphone, without Google Maps. I started walking by empty dark street toward directions that were supposed to be a city center. My main goal was to find a taxi. And luckily I found one. The bad thing is that the driver didn't speak English. Neither he understands the word hotel. So here I am looking for some people who can actually help me communicate with the taxi driver. Luckily I found a young couple. They explained to the taxi driver what I actually want. And it was to get to the closest hotel. So in 20 minutes I am in the hotel. I rent a room. First thing I do, I connect to wireless network. After it I get a bunch of messages from Cid and Marin. It appears that Airbnb address changed last minute. And they tried to inform me upfront. But because I was already in the plane, I didn't know it. And they told me a new address and asked me to come to them. But I said no. I am staying in the hotel. Enough of adventure for me. So instead Marin and Karin took a taxi and drive to hotel. Karin stayed in the car. So the driver does not leave. And Marin was trying to convince me to come with him. Finally I agreed and by probably 1 a.m. we finally all met in person in the Airbnb place. The rest of the trip went fine. But it was my first time when we actually met in person. Time to change. So year passed. We grew into a team of 8 people. We tried to make money in self-hosted market by selling support and consultancy. I think it didn't work well for us. At certain point we introduced Enterprise Edition. We added it as an extra option to support. But it quickly became our main selling point. We wanted GitLab to be a leader in self-hosted market. So we were really surprised when we lost our first customer. It appears to be the higher manager of that company didn't know about GitLab at all. So didn't consider it as an option when he bought a software for the whole company. At this point we realized that we have a problem with awareness. And we need to do something with it. It's Y Combinator. It was a different story, own story. But he will tell you how radically changed the company for us. Thank you. You might be wondering why I'm wearing a lab coat. When I started GitLab.com I wanted to make sure that people would remember the presentation. So I figured why not press as a laboratorian. And it worked. I recently got someone that came back from Amsterdam.rb and said look when you were dressed like that and presenting GitLab.com you were ready to step market was already divided up. I didn't think GitLab would stand a chance. Now this same person is sending an application. So that was great news. I want to make sure you remember this presentation too. So I got it out of the closet again. And it has all the different events I went to in the beginning of GitLab. Now I want to change to the next slide. I have to press two times to do that. We've made a lot of progress since leaving YC. One of the most important things changing our logo one gave people nightmares. And we don't want to be responsible for that. And actually that change happened on demo day after the day we graduated from YC and we met the old contributor. Another thing that happened is that we grew our team by a lot of people and we'll keep growing. And it's really encouraging that the statistics are now saying that two thirds of the people that self host Git are using GitLab. So most enterprises self host so most enterprises are using GitLab currently. Another encouraging thing has been the forced a wave survey. It's the first time we engage with an analyst firm that helps enterprises decide on what to be and we're really really excited that they choose GitLab CI to be the leader in the market. We have the best strategy. We have the best product. And we've shown that we're not just version control. We can do other things. And those things can be the best of breed. They can be the best in the market. That's because we're not doing it. We're just our team members here. We're doing it with a wider community. We have over 100,000 organizations using GitLab and they contribute. Now it's not all roses. We have some challenges. These three things which I'll go into I think are the biggest challenges of GitLab at the moment. One of them is that the clicker doesn't work. GitLab reliability is not where it needs to be. People are a fan of GitLab. They really want to use it but we're letting them down. GitLab.com is not as available as it is as it needs to be and we have to turn that around. And I want to welcome Eric to the stage. Eric is our new vice president of engineering and he's going to explain what's wrong and what he's going to work on with his team to do it. Please welcome Eric. First I just want to say it's been a wonderful month of onboarding. Everybody's been so generous with their time so thank you for that. I feel grateful to be here in Greece to get the chance to meet you all face to face. I can't wait to dive in and start working on solving some hard problems because that's what I like to do. And this is an interesting one. GitLab.com reliability. As Sid said it's not where we want it to be. To quantify that it means we've got about two nines of availability and for companies to build on us and to trust us, that's not good enough. So Sid has put a goal in front of us of make GitLab.com suitable for mission critical workloads and that means we're targeting three nines, 99.95% availability. That's the SLA of our underlying infrastructure providers and we want to peg ourselves to that. And if we reach it, maybe there's a way we can be even better. How we're going to do that is a mixture of people, process and technology. So first in the short term we're going to get prepared to do things that don't scale. At a startup obviously think about automation and exponential growth curves but sometimes you have to do things that don't scale and that's a good thing as long as you have a plan to automate things later. But in the meantime we're going to get our hands dirty. In the longer term we're going to focus on tighter collaboration. We're going to allow the scale of GitLab.com to better inform the features that are being developed for the enterprise product as well as for the community addition. And we're going to fix our infrastructure issues and we're going to adopt some new technologies along the way. So people, again in the short term it's all hands on deck and we're going to use subroute force approaches to improve liability. In the longer term we're going to take our production team today and get them entirely focused on automation. This pie chart kind of represents the time split that would be ideal for them. We want 60% of their time focused on automation, 25% on go forward features that are in the application and as well as 15% on firefighting with issues. You want our build and production teams that are currently separate today to start working together even more closely. Because what they do while the technologies are different is a similar character we think they're going to benefit from sharing techniques and sharing code. And lastly we want the back end teams, the one creating the features to be the ones shipping their code and ultimately pressing the deploy button to GitLab.com. I think I got it working now. I think so. So better process. Again in the short term our new quality department is working on what's going to be our brute force approach to solving liability. In the longer term I want to add a process where we start demoing more. The first tenant of the Agile Manifesto is that working software is the best measure of progress. Can you believe her in this ritual? Are you okay? Okay. Stop with the mic. Testing? Okay. So working software is the best measure of progress. That means I want to start demoing regularly. I promise you the first time we do this we won't like it. But a couple projects in you'll become addicted to this process and you're going to like it. And you've done everything right. It does feel like it's unnecessary but we do it regularly anyway and it helps us solve problems that we didn't even know we were looking for by getting more eyes on the project. Continuous delivery. So this is how we're going to allow the stress that GitLab.com is under due to its traffic to better inform the features that we're delivering. And then we're going to figure out how we collaborate with our new teams. You've got new functional teams for security and for quality and we're going to figure out that working relationship. We've got pictures up here of our new team members, Kathy who's heading up security and BJ who's heading up our quality function. And then lastly the technology. So we're going to be moving GitLab.com to Google Cloud Platform a new infrastructure as a service provider. The top capabilities that we need to start are multiple pristine environments. We also need a continuous delivery pipeline that threads through those environments so we can deploy code early and often and with confidence. The technologies that are dictated by those two capabilities are we're going to containerize everything in our application. We're going to orchestrate those containers with Kubernetes. And then we're going to adopt as much open source components as we can and then write custom tools for what doesn't exist out there and probably do that in Go. And this will be the first iteration and there's a question of can we pack in more into this first iteration? We're going to figure that out this week. But we're going to launch that and then we're going to keep iterating and eventually we'll get towards our ideal infrastructure, our ideal system. So that's it. I look forward to diving into you this week and really crystallizing this project and hopefully leaving the summit with a plan and then kicking it off as soon as we get home. So thanks. Yes, I did. So thank you, Eric. I want to point to another challenge that we have. For the third quarter, we had a great result. We did 90 per cent. The Q4 forecast is for incremental ACV. So the additional recurring revenue we'll generate is going to be 72 per cent of plan. And that's not enough. We want to get to 99. We want to get to 100 per cent. So you can help by creating leads like help if you know a company using Getlap already and form the salespeople by shipping a great EEP product. So these are some issues that our customers really want to see and we have to get them right and we have to get them out in the plant releases and to help bring orders from future quarters. So if sales need your help with a support issue, with a future discussion, please, please do help them. Next slide, please. And the first thing I want to highlight that we absolutely have to get is Getlap Geo. Getlap Geo is the biggest one from our customers. We wanted ourselves to move to Google Cloud. So this is essential to get done. If we don't get it done this year, we'll even have a revenue recognition problem because we promised this feature this year to our customers. So this is the feature we have to ship and every day we can win. Every hour we can win. Every week we can win. It makes it better. So we cannot afford a delay here. I also want to talk a bit about where we're going next year. I want to talk about our product vision, our sales targets, and the big bets we'll be making. So the product vision, I think this user said it really well. Getlap is growing, and we're growing with some of our users. We started as just version control, version control and CI, and we're helping them to go on this journey of continuous delivery. And to tell you more about the product vision, I want to invite on stage our VP of product, Joep. And halfway through the year, we announced this in summer of 2016. We said this is our master plan. Developers should be able to do whatever they need to do inside of Getlap. And then in December, so what are we going to do next? In 2018, we will ship the DevOps lifecycle within Getlap. So this is how it looks nowadays. Have a thing that's called DevOps tools. And what it comes down to, this is much louder. And I think the image here is really applicable, right? Traditional DevOps tools is just trying to duct tape together these developer tools and these ops tools, and often by third party tools. It's not a single experience. I think this is a really good way to see it. DevOps tools is the overlap between dev and ops. And as you know, today, Getlap does a very good job on the left side. We cover the entire circle. So what we're going to do in 2018 is we're going to ship complete DevOps all the way from development to all the way to operations and everything in between. Why are we doing this? Well, when we shipped our master plan, when we included CI, we said if you have a single interface, if you have a single permission model, everything becomes much easier. And the same goes for the complete DevOps cycle. So you don't have to manage multiple applications. You don't have to manage many permissions. You have a single UI and get these unique benefits because everything is in the same place. It makes so much sense. No one else is doing it, but no one else is doing it. And this is why this is so exciting. And I think the best way to describe it is to think about a typical use case. So if you're an engineer and you started an organization, it's very common that you spend the first week or sometimes the first months or sometimes your entire internship, which I've heard before, just setting up the development environment for the stuff you're going to work on. It's painful and it takes a very long time. And what we want to achieve is that, let's say Sally starts the company, she sits down, she opens GitLab, there's the GitLab IDE. She doesn't have to install anything. She just makes her commits right there, creates a merge request, then merge request gets tested. The result of that merge request gets automatically pushed all the way to staging, pre-production, and the results of that are monitored. Everything is auto-scaled. If it's deployed and there are errors, it's automatically rolled back. There's so much power here and there's so much to win here. I'm going to go through the features for a bit, but I think we should talk a little bit about why is this important? And when you look at, when you ask developers and you ask organizations that build software what is important to you? They name a number of things. We have to increase the automation of this whole cycle. We need more cloud-based development tools to paint, to set up everything locally, and you end up wasting a lot of time. You want to be able to actually measure the effect of all the things that you're doing on your customers. It's incredibly important. And lastly, you want to speed up the release cycle time. So this was collected by the same people that called us the leader in CI. And when you look at this and start thinking about it, you think, hey, wait, there's a particular order to this, and in fact it seems that these are dependent on each other. And I'm going to walk you through them. First off, customer experience. I think there's almost nothing more important than customer experience. And it is software today that makes or breaks customer experience. The way I think about it is that when I choose a bank and I recently had to do this, I went for the bank with the best software experience. The best mobile app, the best web interface. Because if you don't have that, it just doesn't feel good. There's nothing more important than this. And if you're a large organization and you want to make sure that this is good, you have to focus on the software. So to be able to create that great customer experience and to ship great software, you have to speed up the release cycle time. Because what happens is that as Wayne Greski says, the puck is moving and you should go to where the puck is going to be and not where it is now. And if you have a very long release cycle time that means that you end up going to the wrong place and the puck has already moved. Your goal has already moved where you want to be. So the shorter the cycle is, the more flexible you can be and the closer you get to that goal and the faster you can get to that goal. So how do you speed up the release cycle time? Well, you do that by automating more. The more you automate, you take each individual step and when you think about how do you automate all these steps in developing, deploying, monitoring, scaling, managing software, and infrastructure, there's only one way nowadays and that's by going cloud native. And this is what we're going to try to do and this is what we're going to do in 2018, all with KidLab. So we go we help you by monitoring your customer experience, we're speeding up your cycle time, we'll help you automate all the steps and we all do it cloud native. All right, let's talk about how we're going to do that. I'm going to walk through the different steps in the cycle and I'm going to talk about some of the things that we're going to do and of course this is not comprehensive we're going to do much more than this things are going to look different but I think this gives us a very good idea of what it's going to look like. So we start with plan one of the coolest things and I think one thing that we all appreciate about something like Google Docs is the ability to collaborate in real time. It's incredibly important that you can quickly iterate on things and you can do this in real time so that you're not in each other's way you're never waiting for someone else you're not blocking in another way. So we're shipping this in issues and merge requests so you can just work together very easily. Next is we're going to start working on portfolio management in other ways. How can you manage multiple projects from a higher level? The first thing we're going to ship in this is Epics and we're going to ship it real soon in 10.2 and there's going to be much more following this and I know that this is a very highly requested feature so I'm looking forward to do it. There we go. And then once you have Epics you can plot them over time and you can actually have a look hey look at this, this is an actual roadmap we have now so one of the features that we're going to build is roadmap so you can see what is happening over time. Alright so much for planning so there's a lot of features here I just highlighted two quick ones let's get into create and it's getting more and more excited the further we get into this I think. We're working on a multi file editor today if you want to make changes in the interface you can only do one file and then you have to commit this file and we already have right now on everyone is running the latest version of kitlab a multi file editor running by any cookie but we're hoping to release it soon. But going from there what is the next step and if you think back about the example that I gave about Sally is an actual web IDE so this means not just being able to edit multiple files at the same time but also being able to have your development environment running entirely in kitlab and seeing a live preview of that so just imagine that all you need to do if you want to edit the website you just press edit and they're an entire environment and you change things you see the update immediately and you can do this on the day that you like five minutes after you opened your new laptop I think this is super exciting this will make everything much easier and then if you have a new project you don't want to think about it we want to remove thresholds we want to remove steps so why do you still have to create a project? why do you even have to bother going to kitlab and push a project so we'll allow you to create a project simply by pushing your new project alright next to verify there's a whole lot in here as well and this is one of the areas that we've been expanding rapidly and I think we've proven that we're very good at but we can definitely add more so we already have in kitlab the ability to see your code quality and what we'll add to this is the ability to check for licenses for enterprises it's incredibly important to make sure that whatever you have in your repository doesn't contain a license that is not allowed you can get in very big problems you want to catch this to the earliest moment and there's no earlier moment than the moment you actually create a merge request alright next you have CI right so you're testing your project you have made sure that the licenses are correct and that the code quality doesn't degrade well what is what remains you want to make sure your application is performant it's fast and it doesn't get slower when you introduce new changes and you want to make sure that it's secure something that I think we're all very aware of today so with the actual change or before it's even live in production or even in staging we'll give you an insight of how your change is affected the performance of your application and even the security and if you have flaky tests which can happen we'll highlight them for you so you can pay some more attention to them alright let's talk about packaging this is one of those things that I think many people are going to love there's many organizations today that work with large binary files be it to deploy their projects or to ship for instance java organizations that use a lot of java they have a big need for something like binary repositories so it will allow you to have a binary repository in each project inside of GitLab and when we're thinking about releasing and I think this is where it gets really exciting we're building on top of Kubernetes and building on top of Kubernetes means we can do these amazing things in relationship to clusters now we'll start talking about that nonstop from now on and I think the first one is the idea of an incremental rollout right so traditionally if you even have set up a system like this where you do an incremental rollout it happens somewhere on the terminal of someone's computer or somewhere on the server and what we're going to do in GitLab we're just going to show you in the application how your changes are being rolled out and then the cool thing here is is that if the error rate is too high we're just going to roll it back and we're going to do that all automatically and visualized of course inside of GitLab so just imagine that right you make a mistake it somehow ends up in production the error rate is too high it just rolls it back automatically and it notifies you I think that's amazing it's so much work that is taken away from you by GitLab we're just going to show you anything automatic rollback so let's say you have a merger request that was merged and that introduced over that raised the error rate we're just going to show you right there where you made the change that your change increased the error rate so you know what exactly which piece of code made your application less stable and actually cost GitLab to say hey we should roll this back and then if we're talking about all of this where does your code go right so this should be your first class support for cluster so right from GitLab you'll be able to create your own clusters and you'll be able to have your own Kubernetes cluster where then your review apps will run your CI will run, your staging, pre-production production environments will run no more jumping into off-baked web apps that are made only to be seen by whoever knows very well about it no everyone can now just look into GitLab and say what kind of cluster we have running and this is how we can manage it and in fact if you have clusters you want to be able to monitor them so we're going to add cluster monitoring so that you have your cluster and if something goes wrong or you suspect something is wrong you can just see it right there in GitLab it's an incredibly powerful idea and I think this is one of those places where we're bringing the whole operation site that normally lives in like a black box at least from the perspective of the developer it's now right there in GitLab for anyone to see, for anyone to have access to you don't have to ask your colleague anymore to say something wrong with the cluster no, it's right there and then now that you're releasing very often right, because this is one of the things that we want to do, you want to speed it up because we automate more we should give you a model to easily release and merge many things at the time and to make sure that whatever you're merging is actually the right thing so by allowing you to test the result of the merge we have a new problem and the problem is that you get merge requests that constantly have to rerun and you would put a lot of resources and release trying to solve this problem now I'm not going to go into the weeds of explaining this but it's very awesome clicker there we go and then when you have all of this there's a lot of cool things for instance if you look at your project you can see all your pipelines you can see what is running what particular change is running in what particular environment you can even pick one up and drag them to another environment just to deploy it so just like moving a cart on issue boards from one stage to the next from one label to the next you can now just say this particular change I now want to see in pre-production because it works fine in staging if you have your application running and a lot of traffic is coming in you want to scale it up well we have the connection with Kubernetes we have your cluster running there it's being monitored there so from GitLab you'll be able to say I want to scale my cluster up I want to scale my application up and use more containers for instance and we can tell you to do that in GitLab but most of us spend half of our day in Slack so why not just allow you to do that through Slack but most of us are ready to allow you to create issues from chat and to look at particular issues and we'll go much further than this and say you can now manage most of your cluster of almost any of your operations straight from chat and chat can report back to you and say this is what is going on in your clusters and in your applications you really want to click alright, last step monitoring so as you know we've been working for a very long time on monitoring in particular with Prometheus and we can do some cool things we can show you application metrics but what we can't do yet is show you how a particular request goes through your entire application stack and with tracing we can finally do that so what you can do is you can instrument your application so that you can see exactly where a lot of time is spent so you can find for instance performance problems a very powerful idea and just bringing this together it's like the future barely anyone is doing tracing let alone building it into their applications and then you now are monitoring a whole bunch of things if today you deploy using auto defops you already start automatically monitoring a whole bunch of statistics about your application, about the containers that you have running but what if something goes wrong? should you always be staring at it? No, of course not so using Prometheus as alerts what we can do is that if something looks strange when you have an aberration in your signal we can automatically alert you and say hey there's something wrong here and you'll be able to go into it and look of course from the same place hey what was the code that was changed here what affected this and then from a higher level you'll be able to see not only how your developers are committing you can not only see how your projects are on issues but you can see how your different applications are doing straight from GitLab, straight from the place where you build all the things now you can see how all your applications are performing and how well they're running or where one of them is done and then one thing that it fascinates me is that there's a whole market of logging products there's so many people trying to solve this problem of logging and people like to move between applications because nothing ever fits but with GitLab we'll just give it to you out of the box. Don't think anymore about your logging applications we are connected to your cluster, we know about your application now you have direct access to your logs as well and then well no really yeah please move it I press the same button and this is like literally my favorite feature so I it's really disappointing that now we have to wait okay there it is alright loading, thank you very much and so I started talking about customer experience, how important is it to modern organizations to focus on customer experience and to make sure that that is good and how you can achieve that through automating and shortening the cycle and you do that through adopting cloud and we're starting to build GitLab around that idea but how do you actually know whether customer experience is being improved by monitoring it and why not monitor it in the same place that you have all your other stuff so what we will allow you to do is to send your business metric straight to GitLab using Prometheus so that means that you can see if you make a particular change for instance you change the color of your homepage you can directly see in the merge request that you made that change whether it has affected sales, whether it has affected customer interactions so you directly combine code with the end result and I think that's one of the most powerful things that you can possibly do to me that is the future of shipping software, directly seeing the result of what you did so if you are just a developer you open your laptop for the first time you make your commit and within 5 minutes you can see how your first change affects the performance of the business I mean that's the future and we're doing it 2018 so using GitLab we will help organizations worldwide adopt complete DevOps and help them achieve a higher customer satisfaction and I think CID is going to tell next about what we're going to do after that thank you thank you Joep for presenting that I want to talk about our other goals for next year and it's not just product we also want to keep growing we're going to do really well this year, we did really well the 2 years before but we have to keep that going our compound annual growth rate our average growth rate has been 178% that means that year over year we're becoming 2.7 times larger in annual recurring revenue and we want to keep that going now the other thing that we're going to do in 2018 are some big bets and big bets are things that align with our vision but that will take us into new markets and get the next slide please we're going to do 4 big bets the first one is a security scan the security scan means that we now have GitLab with auto DevOps, it builds your code it tests your code make sure your code quality is okay it's one missing thing my code is secure so we're going to bundle open source and closed source tools into GitLab that will automatically test your code in GitLab the second bet is integrated DevOps we've seen it today we're going from just dev to ops and we're going to start 2 new teams a packaging team and a configuration team the other 2 bets are even more ambitious we're going to do we're considering doing a peer education company we want to make sure that everyone can contribute right now there are 6 million developers in the world there could be many more but there's very few people who can spend the 3 years it takes to become a developer can live without income during that time so we're going to see whether we can augment their income during that time while they're learning the fourth one is complete bizops we're revolutionizing the DevOps life cycle but a company doesn't need to only create a product it also needs to sell it and with marketing and sales there's the same problem as with GitLab as in the market before GitLab came there there's a lot of tools around there but it's your work to integrate all of them we think we can do better we think we can give an open source integration out of the box between the tools to load data to visualize them and to take action on them so moving beyond 2018 to 2020 what's on the horizon 2020 is the date we want an IPO but there's some other things that I think we should realize and I want to start with the real skills vision I went to a real skills event a couple of months ago and that's teaching people how to learn to program and on the day we we had the assignment on a Saturday a student came to me with a windows laptop and said excuse me for bringing this because she was told that her windows laptop caused all kinds of problems learning Ruby and I think that's not the world we should live in it shouldn't be about what kind of machine you have it shouldn't be about you have to get a Mac book otherwise you can't program next slide please it's not only about windows versus Mac or Linux it's also about PC versus mobile people are moving to tablets people are moving to mobile phones I'm very encouraged that our web IDE will be available on tablets and maybe in the future mobile as well now beyond 2020 I want to remind everyone what our vision and mission are and to kick that off I want to welcome on stage Barbie our chief culture officer Barbie welcome to the stage hi everybody it's great to be here for those of you I haven't met yet it's great to meet you here on stage at the GitLab summit for 2017 so I do think that great people make a great company but I also think that what's great about GitLab is not only do we have an amazing product and enable software development across the world wherever you are at we're able to do our jobs from wherever we are at and that is one of the primary reasons that I actually joined GitLab I was excited about the technology I was excited about what we're doing and what we're enabling in this world and in the workplace but I was also excited to be able to do it from my homes while you're at your homes or the beach or anywhere with internet connection and that's really freed me up and I've got to tell you in the short time that I have been at GitLab my son's math scores have already gone up 30% because I can help him with his homework now never could do that before but I'm happy to be here and what I also think it does it enables us to get the very best people with no limitations on where you sit or how long your commute might be or whether or not you're in a community that's a high tech hub we really can bring the best of high tech and the best of software development to any place on the planet that has internet connection and it enables us to find the best people and to have a successful career you don't have to move from Berlin or Tel Aviv or Austin or the Silicon Valley or Seattle or Boston you can do it from where you're at and that thrills me next slide because I really do fundamentally believe like I said that great people equal a great product and you are the reason we have a great product so what's next how do we build an even more robust product how do we bring companies well we do it by expanding our workforce and how do we do that again through you for the people that we already have here today so I want to make this a little bit of a call to arms in that we all need to find the best people that we know and if they are great for GitLab we need to refer them to GitLab we need to encourage them to consider GitLab and we need to do it for people who are diverse they might not look just like us they might look very different from us but if we feel set then we need to encourage them to be here and one of the ways that I hope to do this in 2018 for GitLab is by bringing an internship program to the company that we aren't a huge company and I don't want to be unbalanced with experienced talent versus very very very junior talent but I do think that the talent we have here can really help to grow both GitLab's adoption from students in universities that then go into careers but they can also help you perform your jobs by young talent who can contribute to the company so you will see that begin to kick off in 2018 as we begin to build an internship program here and so I'll look for you to give me advice on that what are the best universities in your town in your country I might know a lot about universities in America but I don't know as much about universities across the world so as you have recommendations to us about the universities we should be partnering with I want your advice on that too so I'm on Slack you can let me know that that's really one of the new things that will be happening at GitLab this year and I want all of you to understand that we all owe it to each other to help each other be better and that comes from being amazing colleagues and doing amazing work and we can only get there by adding more amazing people so I'm really glad to have you on the team with me in order to do that so with that I will hand it back over to Sid Thanks Barbie Thank you Barbie for that I want to highlight another thing and people have said like I can't believe we lost the deal to Community Edition that's that will keep happening Community Edition will always be our biggest competitor so what we should focus on is not on making sure that GitLab C won't be our biggest competitor it will be because we want to have an awesome open source product but we should focus on making sure there is no other competitors so instead of focusing on the first line here we should focus on all the other lines and make sure we never lose a deal to any non-GitLab competitor out there and I think maybe we should even change this slide and not say CE will always be our biggest competitor but CE should be the only product we lose deals to I think that should be our message I want to say something that I didn't invent it's from Amazon it's from Jeff Bezos and he reminds people that it's still they won and he had three things that I think were really insightful the first thing is you don't ship the process so what we should care about at GitLab is we should care about our users we should care about the contributors and the community we should care about our customers that's it we should never care about the process so if you did the wrong thing but followed the process if it ends up hurting a customer or a user that's a problem if you didn't follow the process but it ended up being great that is great that's what the process is and I love them because they make us more effective more efficient you don't have to ask anyone you know what the process is it prevents mistake that is why they're there they're not there as a defense for doing the wrong thing they're not there to hinder you if you want to do the right thing feel free to change them feel free to go around them it's about the end result the second thing is to embrace external trends we should not fight as a tailwind and I think we did that last year when we said look Kubernetes it's the biggest thing it's a revolution from virtual machines to container schedulers and we're using that as a tailwind we went all in last summit and we are all in right now I think that's a major advantage and the third thing is high velocity decision making the progress of our company you can measure it by revenue growth you can also measure it by how many decisions we take the faster we take decisions the faster we will get ahead just like that release cycle is important you get the code out quickly it's also very important to make decisions quickly and many decisions are reversible so you can take it and you can see what happens and you can reverse them if needed so we should always optimize for speed and the next step is to make sure that we have a consensus company so you should hear everyone everyone should you should let them know you heard them but in the end the person who is the owner takes the decision and moves forward you don't have to wait for everyone to be aligned I'm very glad that tomorrow will be hopefully not using my laptop a piece from our strategy our mission is everyone can contribute how can we in what ways can they do that so we want them to contribute to digital products and GitLab is reinventing the DevOps life cycle we want them to contribute with GitLab so we should make sure that GitLab is available as it can be we should have a great CE version we should have a great free tier on GitLab.com and we want people to be able to contribute to GitLab over 1,800 people have done that and the last thing is we want people to be able to contribute to our organization and I'm always very encouraged when I see people sending in merge requests to our handbook this is the sequence we define where we came out of YC and it's still the sequence we're sticking to today our first goal was to become the most popular on-premise solution for the software development life cycle we did that, we're there it's two thirds of the organizations are using us thank you now we want to become the most we want to generate the most revenue from that and we're very far from that we have a long way to catch up but every year we do that 2.7x we're going to get closer after that and partly in parallel we'll become the most popular SaaS for private repos we need to fix the availability of Gitland.com we already fixed mostly the speed of Gitland.com and I'm very sure we can get there and become the best place because it's the best product and only after we've overtaken SaaS with private repos we'll have a go at the public repository kithub.com is going to stay for a long time the place where people publish public software but if we get the revenue we can make a better product and our big goal is to not just do software it's to allow collaboration to make a great collaboration tool for all the knowledge workers so we can change our culture only to read write now last year at the summit I did a challenge I said if we can get ideas of production working with Google Container Engine I will dance the SID shuffle this year I want to present a new challenge to three different things so that everyone can contribute and not just the developers I want to challenge us to replicate a backup of older.com data to GEO I want to challenge us to ship an alpha version of the web IDE including the editor a web terminal and a preview screen and I want to challenge the marketing and sales people in training videos of objection handling a customer having an objection and a sales person responding to it if we get one out of three me and Demetri will sing the GitLab song we get two out of three me and Demetri will sing the GitLab song while dancing the shiutake if we get three out of three me and Karen I will sing the GitLab song dancing my wedding choreography with Karen I really hope you don't get all three and Karen totally agrees with me this was the presentation we now do Q&A so you can tweet at us but you can also ask them on the YouTube live stream first we will start with a question from the audience no questions from the audience are there questions from YouTube no questions we got a lead I appreciate both questions and leads but I think you appreciate leads even slightly more well if that's the case I want to thank everyone for watching but now we'll keep saying on air and we'll do a live evaluation of what went right and wrong during this practice over talk so maybe you have all the presenters up here and then I'm not sure if Alex wants to join us Alex do you want to join us come on up can we I'll turn off mine and we can pass the mic let's pretend we have a Q&A mic in the audience I'll mute myself now there it is Barbie encourages Barbie encourages to be publicly vulnerable during our executive offsite so cool one mic is for the audience to give suggestions my first question I'm not sure who up there is the right person to answer it but those people out there on the stream watching they may not see everything that's going to be presented in the next few days when will it appear online so it could be on demand the video or the deck the video will be posted probably within 24 hours after the event the team will make sure the quality of the video is good for everybody to be able to watch the replay the slides that were shared today will be posted to the viewers and then the slides for tomorrow will be part of the blog post of the recap which will include the video as well time will be posted to the slide so I don't have the exact time so I have to defer on that also just to piggyback on that with YouTube live you do have a DVR function let's say you just hopped onto the live stream and you missed something that just happened you can always scroll backwards and pick that up I think we can do better so one is the audio so at the beginning of the session there was an open mic I think it was in the back of the room so it was really hard to hear the speakers because you were hearing the shuffling of equipment as well as anybody speaking in the background so we need to make sure that only the speakers mic is hot when we go live on a similar note there's going to be a lot more people here tomorrow we are going to want to post something outside that says turn your cell phones off because we just don't need that happening here during the actual presentation we'll make sure to we've got the mics on everyone that will be stuck to them so it's going to be level to their the loudness of their voice so that should be leveled out by the end of the evening we'll make sure to mention before the presentation starts that people should turn off their phones and we're going to be talking about either unmuting microphones before going on or off the stage or actually turning them off I think their preference on our side is to mute and unmute so it'll be turn on before they walk up on the stage but I think it's good to talk with Alex and the audio technicians to nail that down to an exact science maybe so tomorrow we're still coming up the plan with our awesome audio team but basically I think the process that works really well is before you're going on stage or before you're speaking on stage you'll kind of do a check in and so at that point we'll turn you on and you'll know and with regards to the earlier audio with hearing an open mic in the beginning in the back of the room we were aware and let us know tomorrow if there's anything like that but there shouldn't be but yeah we'll basically we'll check you before you are about to speak when you come off the stage we'll do the same thing and then make sure that you're back on mute or we grab in the mic pack from you I would say three things on the transition make them much faster so hopefully we're right on the side here second it can be longer so energy or fluctuation in the voice so if one person is talking for a long time sometimes it becomes very monotone third thing is when you move around our eyes change and so I have not just stand still but kind of move around engage your audience to the left engage your audience to the right barbie you've made really good eye contact so I would do more of that it's easy to do because I have no eyes exactly use the stage use the stage but try not to do this I was thinking of putting up some tape where people can stand and not stand so they won't be in the camera view going up to the slide so you won't be interfering with what people are reading on the screen or the wall in this case make sure that you are aware but when Yop was on the stage earlier in the back you could see the camera feet from that one in the pathway you could actually see Yop's feet so we might want to not do bright purple tape for example or maybe we could even if you put it just barely right here we would still get an indication so that could be effective I think it also mainly depends on how far you go forward so just people would be better just to put it on the floor exactly because the further you step towards the audience the more you'll be in view of people being on the corner for example so what people usually do in this circumstance is they put an actual line with two end points so you're controlled both left and right and forward and back yeah with respect to the monotone and things like that Yop did a great job very engaging and you were the most interesting to listen to because of the the animation that did that was a very good job on that there was something else too possibly I don't know if the batteries have been changed maybe that would be something that simple okay oh great let's use an old school laptop but I have an HDMI splitter first and second I know it's really old school and everybody likes to have the power of the clicker in their hand but the best and most reliable way to change slides is simply to click the arrow on the physical laptop and it works every time it's not bluetooth which can sometimes fail especially when there's a lot of other transmissions going around the room so that is a failsafe thing and if somebody's manning that presentation laptop they can just simply hit next when it's time let's try the old laptop with the splitter and then have Ernst on standby to hovering over the thing so a few other you want to go first? first for example, use the remote and press and nothing happens Ernst can just press a button usually people have other pre-programmed cues a gesture they hand or something like that that is recognized it does look what would be the word distracting when somebody keeps messing with the remote so if there's some kind of a cue maybe you look back at the slide or something in a certain way that goes to the next slide and that's a small hint for you all who had a great presentation but don't look back look on the monitor here I can't read the letters glasses they're light gray they're tiny okay keep looking at that thing people on YouTube were complaining they couldn't see the complete slide yeah that's something that we're going to be working out overnight but they'll hopefully be very impressed with what they're looking at tomorrow so we can bring those in we'll do the box in box and it looks simple but it's impossible to focus a camera for a projection and a human real life at the same time if you have a camera like that please let us know we'd like to buy them one more thing about one more thing about the transition when the mic or the baton goes to the next person I normally see people reacting more like a team the transition happening off stage is not usually common but whoever is on stage the other person comes all the way up thanks for coming up here welcome it looks more interesting on camera very good apple style yeah yeah chat you can practice last person here's Dimitri any more questions or comments out here any more suggestions the only other comment is who you're talking to this is the person in chat everyone on youtube this is me speaking sometimes you switch between the order which is the team or whether it's the wider audience so just bear that in mind when you're talking about the product and how good and great and good we are so you're saying that we should do more of that we should direct ourselves towards the internet sometimes specifically I don't think it really matters you make it clear that you're talking about the team or you're talking about the product the customer base the community rather than just the team if that makes sense cool if there's no other things we're going to go to dinner possibly with a camera no mic drops we had those already thanks everyone thanks Alex for coming up thanks chat for the awesome introduction this was a great run too thanks my check this mic's hot thanks for watching sorry for taking the stream down what we're going to try now is to they're going to try to make a mobile camera ready but we're still inserting sim cards in the thing and hopefully they can still catch us at dinner but it might be too late so we're going to give it a try for the next 70 minutes until that time it's going to be the super boring stream of this room sorry about that from tomorrow on we're going to try to be a better job and be on time and stream from nine till nine thanks for joining awesome, that went out to two viewers digital just for the presenters to know I'm going to have an extra mic for safety reasons let's hope we don't need it tomorrow right next to you on the table you won't lose any time for me to bring you another one if something happens but let's hope the audio will be like this half on the presentation thank you