 Can you hear me? I think we're a pretty small room, so if people want to come forward, let's just come forward because I don't have too many quotes like the last presenter did. What I'm basically going to talk about is the process and feel free to interrupt me. I have a tendency to speak too fast to catch up with the train of thought, so if you think I'm going too fast, just feel free to stop me. How many of you know about Hasgeek or have attended any of our conferences? Wow, okay. All right, so for those of you who do not know, I run a company named Hasgeek. We organize technology conferences for developers. Currently, so we're not an event management company. We run as a media company where each of these brands, except for a couple of them, are our own brands and the ones that are not our own brands, at least from other organizations that run these brands. Whatever we're doing today's presentation is essentially to walk you through this process of how we've actually kind of gone ahead and taken the truth of building communities ground up by organizing conferences. A lot of times people ask us saying, like, you know, this is not a scalable method. How are you going to like, you know, how are you going to scale? Organizing conferences is hard. We're maintaining communities even harder. And when Naresh first approached me, he said, you know, you guys have been doing it for the last five and a half years. You really need to come and talk about how you've done and what are the lessons you've learned. So this is really going to be a sort of a talk explaining how we've just done this ground up and what are the kinds of things that we focused on and what are our constant learning and challenges. What I will also try to highlight in this talk is how we've tried to iterate. And I think the model that we adopted has really helped us to kind of learn fast and iterate I think to the extent of saying, we will not do this in the next constant or we will not do the same thing next year again. So let me start by giving you a context of why we started Hasgeek. So let me give you a context of why we started Hasgeek. Now the name itself suggests Hasgeek, which means we have geeks. And what we do is we focus primarily on software developers who are working in various areas of web development, back end technologies, mobile. And our goal is really to be able to help them learn from their very peers because our understanding is that technology is changing so fast. But the only way in which developers can catch up and learn is by knowing what their own peers experiences are in their own field. Now back in 1990, between 1992, 2010, one of the challenges was where the software developers really go to kind of like, you know, have conversations. Where do they learn from what is happening in the industry? How do they keep up with practice? Now it's not like there were no events happening for developers earlier. There was Fast.in, which was probably the only conference happening, which was focused on trying to keep content independent and was focused on open source software. See, other conferences that were happening around that time were largely talking about, you know, they were largely focused on giving gyan to developers from a very marketing perspective. So you'd often have people from IBM, Microsoft, et cetera doing talks, and those were not necessarily very practice based talks. They were rather marketing pitches. Around 2000 also is when BarCamp came into the scene where the idea was, let's do an unconference, let's get people to come together and converge around talking about interest. So if some of you attended BarCamp, you know that the format is that the schedule is never announced. It's sort of created just on the day of the conference. And what tends to happen is that people gravitate around those kind of talks or sessions where they have an interest in having a conversation. So let's take, for example, BarCamp, Mumbai, where one of the sessions was literally how do you set up a bar and how do you mix drinks? People who are literally interested in that session sort of gravitated and went to have that conversation. So what our realization was that, number one, developers need a place to go to and to talk about their day-to-day practice. Secondly, by the time that the HTML5 boilerplate really came out in 2010, the issue was where do developers go to talk and learn about technology because also by that time technology started moving so fast in terms of the web that if you weren't practicing for six months, you were literally outdated. So you need to constantly kind of have these conversations and know where tech is moving itself. Thirdly, what we wanted to do was to capture institutional knowledge and create a professional organization where you can run conferences on a regular basis without having to fall back on the kind of learnings. So essentially one of the issues that used to happen with volunteer-driven events was that because a lot of these things were organized by volunteers, the institutional knowledge would keep getting lost every time you did a second edition of an event. So for instance, you did BarCamp 1 and then the next set of BarCamp 2 was sort of taken over by a bunch of volunteers who would not necessarily have the context of the kind of discussions that went in organizing session one. So you ended up having the same discussions over and over again. So our goal was what can we create a professional organization that is driven towards sustaining it and that will sort of capture this institutional knowledge and which will make it faster to do things organizationally. And I think the final thing that we were sort of very focused on is that while there was a huge hype around that time because blogs were in place and social media was happening, what we felt was that you still need physical events to help a community congregate. While we keep saying that people need to, while we can say that people are congregating around Facebook communities right now or they're talking on Twitter, one of the things that is important in the life of a community is that they need physical events to sort of come together, even if this is just once in a while to kind of energize and to have those kind of conversations. So even if you, for instance, take global examples of the chaos communication camp, which happens once in every four years in Berlin, people always come together to this event to re-continue the conversations that they had four years ago or to meet friends. So community is essentially, if you're saying that a community is about conversation, about networks, about collaborations, then people do need a physical event once every while to be able to come together. So with that conviction, we said, okay, we will start a company where our goal will be to keep producing focused conferences that will cater to the needs of very specific developers. So while I've been sort of hopping on this whole idea that our goal was to sort of facilitate communities, things that I'd like to make very clear is we're not the owner of these communities, which means that our goal is to help people come together. And then while they come together in an event, they sort of have collaborations, they have conversations, and various things sort of happen. We've never said that we own any of these communities. We may organize meetups or we may sort of support other groups to organize meetups and there might be initiatives that get spinned off. So for instance, Bangalore JS, which is one of the JavaScript meetup groups, came out of the efforts of one of our conferences, which is called JSFoo, where some of the people who were attending the conference decided over two years that they wanted to get a meetup group going. And then our goal was to just help to facilitate them in the first few meetups and then the meetups have been running by themselves. So that was really what our goal was that we will not say that we own a data science community in India or we own a DevOps community in India, but that we will facilitate these people to come together to make these conversations. And we can facilitate this by two means. One is, of course, by organizing the conference itself and then to see how are the other ways in which these conversations can happen. What we were also very clear about over a period of time was that you need to organize conferences and build communities around emerging problem areas and not necessarily around the technology itself. Technologies come and die. I mean, let's take JavaScript, for example. You have new frameworks coming up every now and then. Now, the challenge really over here is do we then sort of center the conversation around the framework or do we then sort of look at what are the larger problems that are there? So for instance, do we focus then on developer efficiency instead of saying AngularJS is better than React? So one of the things that we also decided early on was that we try to avoid as far as possible to focus communities around a particular technology but rather look at a broader problem area. So data science, for example, the conference that we run every year called the Fifth Elephant just focused on a much broader problem of data science and is agnostic in terms of whether you use Java or whether you use Spark or anything. What we're doing over here is trying to address a bigger problem of how do you mine data at scale? How do you mine data with speed? And how do you mine intelligence out of data? Now, again, like I said, the question that we are often bombarded with is like, you know, why do you kind of have to do this hard work of organizing conferences? And I will tell you, as somebody who organizes practically eight conferences a year, it's not fun. It's a lot of, you know, getting a lot of people to collaborate and oftentimes I'm like, you know, I'm on the brink of burnout or like my hair will just fall off my head. But at the same time, this approach is critical. And the reason it is critical is because I think like the previous speaker sort of mentioned over here is that technology enables certain parts of scaling a community. At the same time, it's important to be able to do things that hold the community together from time to time. And if that means hard work, it does mean hard work. Again, people often tell us in conferences, don't scale. At the most, you have 5,000 people or 10,000 people, you know, you're not going to reach millions of people. And like I said, you know, we're organizing like five to eight conferences here that, you know, someone mentioned to me earlier last year saying that you must be out of your mind. So what I did mention in the beginning was that we're not an event management company, which means that we will not organize an event for someone else or that, you know, we're not organizing these conferences as, yeah. Because that is the first way. Now you can scale a community. You can look at the ground up method to look at what aspects of the community scale, which is what I will talk about subsequently. And, you know, those aspects can be enabled by technology or they could be enabled by other sources. And that is what we have learned from our experience that what you really need to do is to kind of do this hard work to at least have the first layer set up. Let's take infrastructure, for example. Like, let's take FRIP Cuts instance for instance. Why did they have to go and build a courier company in the first place? Because that infrastructure never existed in the first place. Similarly, for our conferences, we now invest so much energy in getting Wi-Fi set up literally at the venue so that even if you're not using the venue around the year, Wi-Fi is still available at that venue. Why do you need to do that? Because if you don't have that layer first, what are you going to build on top of? And that is the approach that we've taken over here also saying that yes, it's hard work. You have to do this. And then from here, you start to look at, you know, what aspects do you productize or what aspects do you then use technology to scale. And I'll also explain to you how we've done some of that. I think the point over here is also that, you know, we're not just saying that we will do this grunt work of organizing conferences only. One of the things that we actively do within the company is also build a lot of software. And some of the software is intended to either automate certain processes that are involved with organizing the events or this intended to kind of accentuate the kind of conversations that happen. So to give you an example, one of the key aspects of our event is the conference badge. Right now you have this badge. What we've done over the last five years is literally I created every producer better badge. And the reason why the badge is so central to our conference is because first we do not sell sponsor database to the sponsor as part of sponsorship. The sponsor and participant can voluntarily exchange information at the venue itself. Now to enable that, how do you make it happen? You can do, maybe you can do the standard thing of exchanging visiting cards. But what we sort of realized is that technology at that point in time enabled those transactions to happen in a sort of smoother way. So what we did initially was we actually got NFC cards and we used to produce badges like this to insert an NFC card inside. We used to set up NFC readers and every sponsor's booth and we built our own technology called contact point where then a participant could come, have a conversation with the sponsor of the booth, tap their badge on the card and the information would get communicated. Whatever information the participant shared would get communicated to the sponsor. So that enabled the sponsor and participant to have a voluntary exchange of information without us having to be a party to that exchange and without us having to violate the participant's privacy to make this exchange happen. Now while we were doing NFC back from 2012 to almost 2014, the challenge was that readers were slow. There was a huge problem with respect to Wi-Fi. So if your internet sort of failed, the NFC readers would fail and then we had to keep iterating by saying that we will actually set up a LAN network to make this happen. Finally, we said leave all of this. NFC cards are also something that people usually take back. They don't return. Then you've got to replenish your supplies and what we started doing right now which is sort of not very visible in this photograph but maybe I can show you a bunch of photographs later is now we actually produce a badge which has a unique QR code on it and that is pre-printed before the participant comes in. So every participant gets a badge with a unique QR code. Sponsor simply has to sort of scan the QR code because QR code scanners are very easy to get or you can even download an app and scan it off and that enables a much more faster transaction of information between them. The other thing is it's not just, I mean the badge is not only central for this particular reason. It's also one of the things we realized early on was that when you give out a printed schedule to participants they usually lose it. Then you put schedules out on the board like this and people are still scrambling. Finally, we said okay, let's do a simple thing. Let's eliminate this plastic pouch because first of all, not environmentally great. Secondly, we now introduce a badge which is a flap and I'm sorry I'm not carrying a sample of the badge but now we literally sort of print the schedule in the badge itself. So all the participant needs to do is to either like turn the badge back and look at the schedule or just open it up. So what I'm essentially trying to say is that what our model has also enabled us to do is to make the experience smoother and at the same time, look at what parts of a conference can be automated and can be made easier to kind of organize over here. So this is just one instance. We've built a bunch of web apps. Right, so I'll cover that now subsequently. So I think the biggest factor that sort of drives in getting all of this together is around what do you have a conversation and so for us, the sort of the guiding principle of a conference is the content. How do you kind of get the content together in such a way that you then enable that conversation to happen or rather even before getting the conversation to happen how do you get that audience in the first place who will then form the community to go forward. So one of the things that we sort of stuck very strongly on is that each conference has a fairly tight focus it has a very tight focus it means it is meant only for a particular group of software developers. So let's take for example the conference that's happening now on 14-15 April. It's called Rootconf and it's meant only for systems engineers and for DevOps and for people who are dealing with IT infrastructure at large scale. We're not interested in having designers or front-end developers to come over here. It's going to be two days of just discussion about how do you design complex systems, how do you cope with failure and how do you kind of scale your infrastructure. Whereas you know the conference subsequently is bringing in only product engineers and product designers because the idea is that we're only going to talk about design around product and how do you kind of help your user to stay in the home of your product. Similarly in July when you organize the conference we'll have a data science conference and that is probably the only conference we do where we try to bring in a slightly more diverse audience because data is not just a technology problem it's a business problem. So you try to get the tech and the business groups to kind of talk to each other because this is where the conversation really needs to happen. We're focused on a very specific kind of crowd. JS4 will bring together only JavaScript developers. You won't have a systems engineer coming in for whatever reason. So let me sort of... I think your train of thought is faster than mine so maybe you can hold on to your train of thought for some time. So I think to kind of to get to that question I think number one was every time we decided that we're going to launch a brand or we're going to do a particular conference we were very very clear that it's going to cater to a certain group of people and also we had to think slightly ahead of time. So for instance when we launched JS4 in October 2011 JavaScript was just about beginning to catch up. It hadn't gained the kind of international attention in India but we said this is going to be a very important language for web development and for web developers and that's why we were the first people to move on that front and that sort of enabled us to stay ahead of the time and build a community. Today there are a number of JavaScript conferences. Similarly when we launched the fifth elephant which is our data science conference in the following year the idea was to be able to bring together a conference where first of all business and tech both have a conversation at the same time you do want to bring in people talking about core concepts of data science without it necessarily becoming too academic. So for once year is a conference of practitioners which even academics will come to because there is a sort of practice going on over here. So every time we launched a brand it was very important to ensure that there was a clear focus and there was a very clear intended target audience in mind. And over a period of time one of the lessons that we learned is while it's great to have a lot of content in a conference so I suppose over here there are four tracks going on simultaneously we actually made an experiment where in one and a half years of organizing conferences we decided every conference will have a single track only which means we will make sure that everybody is sitting in the hall at the same time and this is important for two reasons. One for a speaker this is important to get that kind of an audience and so that the audience is not split about the kind of content that she or he is delivering. Secondly to have a single track means that you're going to keep your conversation focused only around that particular theme. So every year when we organize the conference also there's a very specific theme in mind. So let's take root con for example last year we spoke about virtualization and it was very clear that we were going to talk about virtualization and computer only. This year it is only about learning from failure. How do you design complex systems of failure? So every year the conference also has a theme and we ensure that there is sort of a very tight focus and all the talks are sort of centered around that theme and that you keep your audience and your speakers sort of glued together by saying that you're only going to do a single track or two tracks depending on the kind of content you're covering. So that was an experiment that we did one and a half years ago and of course it was very, I think it was very scary for us also to say you know what happens when you do a single track only because if your talks don't do very well, if even three out of your 11 talks don't do very well the impression that the audience gets goes back with can be pretty embarrassing. But we decided that this was the way to do it and so that was one of the iterations that we sort of did. The other thing is how do you I think a lot of people keep asking us saying how do you get speakers for your conferences and this is something that we've experimented with over a period of time. When we did our first set of conferences between 2010 and early 2011 we literally handpicked speakers and said okay you know these are speakers who are going to talk. But I think subsequently we arrived at these two realizations. Number one we said is that if we want practicing developers to talk to other practicing developers how do we discover who these practicing developers are and where they are sitting around. And secondly how do you kind of what kind of a system do you build to be able to attract more people to come forward and talk. And that's when we built one of our first web applications which is called talk funnel. It's somewhat similar to what convention is. But the idea of talk funnel was that here's the place, here's the form you can fill it up. You can propose a talk to speak whether you will actually make it to the stage or not depends on the kind of process we follow. So what we did initially was that we got people to sort of say that you propose a talk and then you popularize your talk by tweeting about it by getting more and more people to vote for your talk. And so in the first two years all we did was get speakers to submit talk and then build traction around it and then see that whichever talk had the highest votes actually sort of get to go on stage. Of course in one and a half years we realized that the system can be gained quite easily because you can get so many people to kind of vote. So what we did was we introduced another, we introduced a second layer of filtering where we said that what we will do is now get people who are our past speakers or who've done significant work in the community to now come and look at to give like a third perspective to the talk. So while you have the votes let these people also sort of like take a look and say like which of these talks qualifies to go on stage and that's when we introduced the idea of an editorial panel. The editorial panel is basically people who are from the community who've been past speakers and who've done interesting work and for us this panel is extremely important because these are the people who will then go forward to kind of like play an important role in the community. If you have to grow a community you also need to be able to reinforce and hone leadership and that was the way we kind of built the model over here. The editorial panel has also gone through a whole number of iterations while initially this used to be a group of four people who used to come together, select talks and all of that. One of the challenges was that because it's all volunteer driven it was very difficult to get each one of these people at the same time to kind of curate talks and therefore over a period of time we said we will now refine the criteria and the way we kind of do editorial and now we literally have just one or two editors per conference where the idea is that the editors should have the experience of conferences either they speak at other conferences or attend other conferences and they should have a sort of a passion for the community that they really want to build this community and therefore they're giving their time and energy over here. So all our editors are currently all volunteers and we now kind of use this criteria but also as I will sort of point towards the last slide now what we're doing is expanding the whole pool of editorial by saying that they're not just one or two editors per conference but you actually have a pool of people who are interested in curation and this could be curation around the conference or meetups or other forms of content in that collective. I think the editorial model has also been very useful for us to identify a number of people who have an interest in promoting the community and that makes it very interesting and important to kind of propagate the community itself. I think at this point the question would be like you know who sort of speaks and why do they speak? We don't pay our speakers. Many of our speakers literally just travel across the world and pay for themselves and even come here to speak and so like I said every person has to propose a talk to speak and what we then do is to kind of like select the talk I mean we review the talk, we refine the talk, we even give guidelines and everybody has to submit an outline and there's a whole rigorous editorial process which we could talk about later in the Q&A but you know who gets to speak it's a practicing developer who has to speak to other practicing developers. You have to be able to provide an insight to the audience and only then is your talk considered meritorious because if the audience is sort of coming there to hear you speak they need to take back something either they take back in terms of direct correlation to what you're saying or they're able to abstract what you're saying for instance if there's somebody from Flipkart who's sort of speaking at stage it's not necessary that everybody in the audience is also operating at that kind of scale or is dealing with those similar challenges but the speaker should be able to have the ability to present a picture in a way that the audience member can literally abstract and say while I don't operate in this domain or I don't operate at this scale or I should not be applying this solution to my problem. The way a speaker also presents is that you're expected to provide an answer to why did you make a certain technological choice and not just say I made this choice and this is how I did it because how I did it is available by diamond dozen on the internet so our focus increasingly in the last couple of years has been to push speakers and say please talk about conceptual problems please talk about why you did something versus why you did not do something because that choice is very important for the audience member to understand and as much as it's important for the audience member to understand patterns it's also interesting for a lot of audience members to understand what is the antipatterns and what they should not be following a lot of people who choose to speak on our stage do it for a bunch of reasons and I'll try to highlight a couple of them I think one is of course a reputation when you speak over here you're actually sharing your work with your peers and the kind of feedback that you get from them is phenomenal and that is one of the reasons why it drives people to actually come forward and speak over here secondly it's an opportunity to share your work and you're sharing your work directly with the people who are recipients of could be kind of using your work so it's also good feedback over here and a lot of people who have gone on to speak at our conferences then go on to also speak at other conferences which doesn't mean that doesn't happen anywhere else but we've constantly found a lot of our speakers then sort of like going on international stages and speaking and again this is one aspect of this whole process the last half years that we're now trying to focus on even more which is how do you give more and more reputation to your speakers how do you actually scale their visibility and their reputation beyond the conference itself and that's sort of now leading us to think about content in various forms it's not like everything works fantastically well and I think in the life of a conference there are a lot of things that change and as much as a conference sounds very mundane it's almost like an ecological system that sort of transitions year on year now while I said initially that we do need to crowd source content we've also realized there are times when you have to literally infuse content into a conference so again let's take the example of the April conference that's coming up because the theme was so specific saying that we're going to talk about systems, complex systems and how do you design them for failure one of the challenges we realized early on was that the kind of people who do this kind of work are not necessarily sitting around waiting to speak at a conference they're probably like fighting fires every now and then so we literally had to go out and seek people who were working in this domain so we sought speakers who were working with my SQL and said you know talk about performance and failures so that's where the network comes into play which is you either seek feedback from your participants and say you know who do you think is worthwhile to go and approach and then we also keep an eye around to see who's actually doing interesting work and what projects are interesting enough for the audience to kind of listen to so for instance we realized that you know if we have to kind of do this kind of discussion around large systems you have to be able to talk about chaos monkeys so we literally sought somebody who on the Netflix team actually worked in chaos monkeys and said you must come here and talk to the audience and tell them how you did it so in a certain way what is very challenging for us always is that how do you keep the content ahead of the times and how do you keep it relevant for the community because that is really how people are going to congregate for having that discussion in the first place it's not again I think the other challenge also is that because you know as people are sort of paying to kind of attend this conference is a huge tendency for people to just become consumers of content so literally they're running from like you know one session to another and say where are the videos I missed this session I need to see it and literally it's become like consumption of content so for us it became important to understand how do you kind of like you know what do you introduce into this dynamics and how do you make them active participants in your conference itself so one of the things that we've noticed at international conferences for instance is that even if you have a even though you have an attendance fee there are people who actually take an active service and they're using certain aspects of the conference so let's take for example the chaos communication camp that happens in Germany every four years people literally travel from all over the world bringing all kinds of crazy things with them to produce at the conference somebody got a soup trailer from like two states away and actually sort of made soup and served it to everybody at the camp itself there was a whole bunch of people who basically got vacuum cleaners and old pipes to reproduce the system of messaging and for two days people were doing literally like you know sending messages to each other through vacuum cleaning pipes now what's magical about that experience is that because you're a producer and not just a consumer the environment completely changes and for us now the challenge is how do you make your participant an active contributor to what you're trying to do so now going forward the question for us is how do you have installations how do you have activities that help people to come out and I think one of the experiments we did last year was to say that you will have talks that are going on at any given point in time what you will also do is have conversations around very structured questions so we introduced the birds of feather sessions and said okay you know people have a choice you can either go and listen to your talks or you can come together and we'll have a conversation around a question so for instance you know one of the questions that we introduced at GroupCon last year was how do you choose your options with virtualization you know each of these and people had a conversation there or we had a conversation around like you know why do you choose MySQL when you choose Postgres and of course people sort of came and had a huge fight in almost like a religious bloodbath but what was interesting was that for once you could actually take people out of that experience and get them to have a conversation outside so for us it's also important to keep experimenting about formats within the format itself in order to have the engagement going on constantly because what matters in the conference is the content it matters what kind of conversation you're having and what is the kind of take away that you have the other challenge of obviously like in the last five and a half years is that when you're always doing like you know newer newer conferences and subsequent editions the very people who were involved initially will feel left out and I think that's not just the case for conferences it's the case for any community I have been a runner and I have been associated with the running groups since 2010 I think two years after I left the group and then joined the group again last year it was very difficult for me to get the context of the group back and participate actively to the extent that I dropped off in the next three weeks I think that's always the challenge for any community in terms of how do you keep old timers engaged and I think that's the challenge that we are also facing and the reason why you do need some of these old timers to come back is because some people literally come into a conference to have conversations with these people so it's almost like every six months you're coming back because you're continuing on a conversation that you left off some time ago so while there's no clear answer to this there are a bunch of experiments that we've tried out that all our old timers sort of get a loyalty pass so we just invite them to come because we think that their presence is very valuable so I think like I said earlier this format has sort of helped us to definitely make changes rather quickly and also to realize where is it that you actually build technology and where do you kind of scale up now I did mention that one of the changes that we did last year was to introduce birds of feather session the other thing that we've also started doing actively is to get audience members to come up and do 500 presentations and oftentimes that's a very interesting format because we find that a lot of these people then go forward and then the subsequent years go on to propose talks and become speakers and that's a good system to have because you do want people to graduate to those levels and you do want their aspirations to rise now as much as content remains very critical to the conference I think the overall is very important for the whole conversations to come through so for instance over the years we've experimented with how with the kind of venues we've used we've started with we've been very clear that you need venues which have comfortable chairs you need armrest because people are going to be sitting through for two whole days for almost 16 hours you do need to give them that kind of comfort so we've started off with like small time venues like Dharmaram college and then gone on to kind of like you know produce our events at Nimhans and MLR conventions etc and with these venues also we've constantly been experimenting to see how do you structure the venue in a way for people to have a good flow and to have a good experience similarly we've invested a lot in producing infrastructure earlier we would actually sort of like in the first couple of years in the first one and a half years we were actually outsourcing video recording until we realized that nobody knows how to actually produce conference videos everybody knows how to produce marriage videos and I don't mean to say this to put down his job but our realization was the kind of videos that we wanted to have were just not happening and so so no so I think video recording has to be viewed in different context I mean one is being able to so even for us the thought process of video recording has gone through a lot of changes in the sense of why do you produce videos in the first place now standard practices you produced for archives how many of you are actually going to go back and see these videos no it's just going to be in the archives our realization in the last one year was that you need to produce a video in real time and actually make it available as soon as the session is over so that if there's a conversation going on around twitter for that particular talk then you can say okay here's the video you can also take a look at the snippets now we don't expect everybody to see a full 45 minute video but this is the way in which you this is one of the ways in which you scale the reputation of a speaker so for instance last year we had a talk by Vishy Malhotra who was then the clear trip and he talked about how clear trip actually built the mobile app as a product and he literally talked about it from a product experience the moment his talk was done we immediately put it up live I mean it was streamed it was recorded and then immediately there were a bunch of people from your story who wrote a large article about it and Vishy got his own reputation as a result of that similarly we found that you know people who sort of made either controversial statements in the video or who've talked about open source projects that are interesting putting up the talk available making the talk available immediately in real time sort of then enables those conversations to continue either on Twitter or through other forums now now it will drop off I mean there is no doubt that it will drop off but some videos for instance again a lot of traction like in 2011 we had a talk on backbone and that video is still very popular in still extremely widely viewed for some reason or the other it's become like this tutorial that people just have to watch and the speaker has also sort of made it very popular now for us the questions obviously as you produce these videos for archival value and you know I mean as our own thought process is changing the question is also like you know what parts of the video can you then abstract and put together and stage it together with other forms of content so I know that the previous speaker was also talking about video but I think the hard challenge with video is how do you get people to consume and I think Buzzfeed has some interesting models around this and there's something worthwhile to learn about in that sense similarly it is I think Starter also has a good use case while Starter is not so focused about the talk that happened in the conference itself they do a lot of short interviews with speakers outside and those gain a lot of traction we've had a sponsor in the past Aerospike who's literally capitalized on popularizing the database through this form of content where what they would do is get the CTO to speak at Starter and then immediately get short videos produced soon after the you know while in the sort of the breakout sessions and then make those videos popular so I think people have sort of been experimenting with video and coming up with newer ideas for us the first challenge was in just getting the infrastructure in the first place so we now literally live stream since 2012 we've been live streaming our talks and I don't mean to say that live stream is more popular I mean sometimes you just have like 5 or 6 people watching the live stream and sometimes there are a lot of people watching the live stream but just the fact of the video being I mean of the video of the talk being available for you from your home is also very important and for us it's been a huge investment in terms of cost and time to get this infrastructure going every now and then we have to replenish and revive this setup sometimes we literally say okay throw away everything we've built together in the last few years and actually get a new setup going and we've now come up with our own form of editing and that goes simultaneously with the kind of changes that we've made with stage production as well we've been trying very hard to look at how international conferences brand and get the setup right because stage setup is not right only for the aesthetics it's also important from the point of view of the speaker's presence on the stage so for instance back in 2013 we would make elaborate visuals that would go on our stage and they'd be so beautiful that people would be watching the visuals instead of watching the speaker and that made it worse with the videos because oftentimes you'd find I remember back in 2013 and we can play that video when Robert Nyman came to speak at JSQ and the theme that here was bots he was literally standing over here and the and the artwork literally had to end and as we should go up behind him because the artwork at that time was all was all sort of themed and graphic we finally came down and said no we look at the experience of other conferences and internationally and sort of like you know tornness down so that the speaker gets as much visibility and at the same time the brand sort of remains consistent so similarly I think the other thing that we've realized is even with respect to simple things like food usually the challenge that you sort of face over here is that people are just sort of served very carb high food and post lunch it's very hard for a speaker to hold the audience's attention and at the same time there's a genuine problem with the whole vendor system particularly in India where you know things are taken for granted over a period of time and the last one and a half years we've said okay we'll go on to introduce food courts where you can actually sort of give a choice to people in terms of the kind of food they eat and that I think is extremely important with the experience that you have at the conference similarly collateral over a period of time we've also realized that what people do want to go back with our gadgets and devices where they can actually experiment after the conference so since last year we've started giving out devices that are sort of relevant to the conference itself so for instance we gave a virtual reality headset at droid con and we also sort of coupled it with a couple of virtual reality workshops that were conducted by common flow and a bunch of other independent trainers and it was a great experience because you know while virtual reality is not entirely mainstream it's just something that people want to play around with and gives them a sense of happiness it's also a way to kind of like you know push your consumer to becoming actually a producer at the conference itself. At root con for instance we said okay the most logical thing to do is to give people UV keys because you know it's all about security and you know the kind of the way in which you protect data so similarly I mean like you know when we do MetaRefresh this year we are producing prototyping kits to enable people to actually prototype so the idea is to also sort of like you know to give an experience to people that goes beyond the conference itself yeah I think we'll have to stop but I think what I quickly try to do is to to literally summarize in terms of how things have sort of been moving forward while I've said that you know we've been producing our own brand of conferences what has also happened with us now is that we realize that in some cases you actively collaborate with others to be able to kind of produce a community and to produce content itself. So Julia Khan is a great example where Viral who's a co-creator of the language and it's not funny I was just here six months ago when functional conference happening and I was having a conversation with Viral right here when we decided to do Julia Khan while you know Viral was a speaker at the Fifth Elephant and Julia started to gain a lot of traction in 2013 as a result of the talk that you did at Fifth Elephant and eventually it came down to a point last year where we said okay you know if we have to do a conference and build the Julia community in India it's going to be a partnership between the foundation and between Haskeek. So this has been a great outcome for us similarly this year we are producing conferences with other partners like IcePrit and others to recognize what are the kind of challenges in the payment space for developers and for businesses and I think the other interesting development that sort of happened since the past two years is how people who actually volunteer at a conference are now becoming our collaborators. So for instance people who sort of presented, who actually did a workshop at our data conference last year have now gone on to become full-fledged instructors in the area of data science and we are now producing the next edition of the Fifth Elephant along with them being curators of installations and the talks and things like that. So I think in some and substance at least what the model has enabled us to do is to kind of help us realize how do you actually kind of grow a community and from here if you want to kind of like diversify into looking at content as a business or you know looking at how do you kind of scale for your speakers and others these are the kind of interesting insights that have come to us over the last few years and so now the next challenge is where do you build products from here. So I would have liked it to be but I would be around so we can still have this. There was a lunch and we can also have some questions over lunch but maybe somebody else could ask a question. Sorry, I always wear an editable hat for my conferences and I'm always like please keep quiet after a point. Right, so if nobody has questions then he can spout but the content is a detail miss and I will not make any bones about the fact that selection is extremely painful and hard for us. It's painful and hard for us because we have to go through so many hoops as much as I think the biggest lesson we've learned in the last couple of years is that as much as we would like as much as great content is important a great speaker is equally important to deliver the content and I say great it doesn't mean like you have to be fabulous and have to have spoken several times but you do need to have the ability to communicate the content which is why we now made it mandatory for people saying if you are a first time speaker or you don't have links to videos of talks you delivered in the past, please in two minutes like do a self recorded video and show us that you can actually deliver this content because I can have great content on stage but if the person doesn't have the ability to communicate it's a problem. At the same time we've also introduced a rehearsals program now before every conference three weeks before we have rehearsals so that you get to kind of refine and fine tune your content and you also get feedback from a sort of semi-live audience and from a mentor who actually helps you to kind of do up your slides and things like that so for us it's very important to kind of make sure that the person who actually goes on to the stage is well prepared and that the person actually gets what the person is out to kind of do on stage which is to get that kind of reputation and for the audience to get that kind of content and in spite of all this we still have hits and misses and so now the thing for us is how do you kind of get smarter and smarter people on to editorial because as much as we might say that machine learning and all will solve the problem eventually there are still problems that technology will not be able to solve so you still need to have people who will be current and updated with content who will be able to kind of make that, will be able to discern that's right there is nothing, yeah so that's where we announced sort of changing the model and saying that the conference now becomes only one of the activities to congregate around a year we now do a whole bunch of things around this thing so we've now started doing a bunch of data science workshops there will be meetups proceeding the conference and meetups after the conference and a whole lot of things, we're also going to be now doing smaller conferences as well as a deep learning conference that we're going to be doing before or after the 5th elephant to sort of focus the conversations better and things like that so now I mean the challenge for us as a business now is to see how do you now keep this community alive and not just sort of focus it on the conference, it's taken us 5 years to get to the logistics and infrastructure of the conference now the challenge is how do you keep this alive yeah we can talk about lunch, I always like listening to ranks it helps me figure out how do we do things better ok so if not then maybe we can have lunch and have