 Thank you all for coming to this meetup. Thanks for staying late. How many of you don't know me? Okay, so I'll go through some organizing this group together with Michael. It's been sponsored by a couple of great companies, Pivotal and PayPal. This is the premise. Beginning of last year I abandoned the job at a financial services company, JP Morgan, and joined this startup, which is Pivotal. Thinking about that I will join, you know, a unicorn club. Fair enough I went to Silicon Valley, got the Kool-Aid from the from the valley and then in the first in the first meeting that I had with a Fortune 50, you know, I tried my best, got it all the technical stuff, everything that they need to do like step-by-step and then halfway through the first ten minutes everyone was blown away and then I told myself okay hold on I need to explain from a level like from zero level like what it's actually all the technology all these buzzwords and like how they can do even the simple stuff so like teach them how to walk before they begin to run and I wanted to make this presentation for a long time continuous delivery and why I think everyone should do it because if you're in software business there's no other way. I mean survival is not mandatory so even if you have a safe job now even if you're like you know you have budgets you have people hundreds thousands it the disruption is here so I will not go into a long conversation why you should do it and the disruption is on the door and like you know you should feel unsafe instead I'm giving you a glimpse where like you you can jump along and go through the salvation so the way I want to do it is this is basically a pipeline that we have for one of our products right so this is concourse this is basically how you have the pipeline from the simple commit to a delivery in production so these two is kind of like you know if you go to the technical folks and tell them okay forget about all your emails forget everything just open this dashboard and it will be immediately telling you where in your where is your product which version it is what's the product here's the legend and this is actually happening live and you have multiple builds I can switch between them and some of them they will they're started or green which is good but the talk will not be about the technicalities it's just like if you look at these tools and everything that has been done it's basically to explain to people why are they doing it and where's the bottleneck basically so that they can actually adopt kata and then improve where it's stacked or it's not working as fast the next part is basically this is cloud.gov you would probably never expect this in your entire life it's the US government adopting Cloud Foundry as their cloud capabilities and this is basically the setup instruction for a new employee in the government basically how can I set up and get working and it fits on a page you don't have to raise you know countless of tickets work for approvals do all the training sessions compliance it's basically get an account install the command line interface verify that it's actually working so that you have this version then login and you're good to go this is your sandbox you can begin improving all the applications under the US government and this is basically adopted after they failed miserably with their health care and this is literally like the setup page and I'll actually double check if so that you it's on the docs cloud.gov so it's authentic now this is my presentation you can follow along you can open in your browser this is the the stuff that I just pushed here's the logs so this is what I pushed a few minutes ago so I really iterated very fastly so that I can change the the way I have my messaging so let's begin continuous delivery why everyone should do it this is my Twitter handle if you want to keep in touch after the meetup or ask me questions the next one I'll basically start with the objectives of this session if you do all of this then you can probably do something else through the entire session because it means you've mastered all the concepts you can do it and everything is cool can you release new features to your customers every week every day can you do you have the confidence to actually onboard a new a new person with the pro on the project and have him commit on his first day and then you know sleep safe confident that okay he didn't broke anything there's not that handoff which is three months six months were like he needs to go with the baby steps until he gets you know access to production or access to any product in a meaningful way that means that the guy within your organization is not able to have a say because for six months he's learning the weeds and probably after the six months he'll find that okay I'm probably in the wrong organization let me like call my recruiter definitely it's something that we should all dream you know keep it somewhere along our core values we're like yeah we cannot do it tomorrow but it's something that we should strive for in a year in two years in five years and basically you know do small steps because this is the everything that we can hear on the market the value of going fast right like competitor a bc can do it faster than us we'd need to do it faster how do we do it okay call you know a bc learn how to do it but ideally is that you cannot put a dollar figure on this fast value like you cannot charge the number of services you cannot charge the number of developers is like it's opportunity cost basically how much do you put a value for a new product that's in it's on the market and you can deliver like two hours two weeks two months before your competitor like what's the opportunity cost you can only guesswork it you can make you convince yourself that you can actually put a figure on front of you but it's not realistic this is the union the unicorn club and this is getting very crowded so from 2011 January to July of 2015 this is all the companies that you have on the marketplace that have evaluation of more than one billion dollar and I mean probably some of them they are you know you use some of them are no and even for myself when I actually opened the Twitter and actually looked at this I was expecting that will be maybe a few maybe like you know I'll have 10 maybe 20 among this lifespan within a year what it actually means this that you have dozens of fortune 500 companies that went out of business because if you have new players what happens to the old players and then you think okay so I'm gonna do this I'm gonna get my developer I'm gonna go really fast I'm gonna do all these concepts but then in your mind you have this strange feeling okay it will all go wrong I'll give you know I'll empower this person I'll give access I'll do all rules by the book and then you know next day I will not have a business because you know they will fail the business they will you know they will fail miserably because you know there's so much complexity that we have in software technology and to that being said yes there is still complexity there are many companies that have tried to overcome complexity by different means and they would you know bought expensive tools paid expensive consultants had style you know hundreds of checklist processors you know we don't want to have failure in production but the key to that is that you cannot actually manage complexity you will start with a small product it will be you know you'll have a team of five it will be around this table everyone will know what to do it but soon enough the the product will grow it will be 20 people 25 people you cannot have you'll have to kick people out of your conference call so that you can actually communicate so it will grow this is just a you know something that it's reality and what you can do is something that you know Facebook Google have done and the way they've done it is the CTO has thought okay what if we can actually optimize for a reversibility so what if we can always do this continuous delivery but if we fail then we can basically have a second of that failure and then we'll basically revert everything very very fast so nobody will notice and then over the cross of a whole day you can have hundreds of failures but if they only take for a millisecond does it matter will someone on your client side actually notice that be sure it's not basically this is something that was coined at Facebook and there is on their tech blog they actually publish the way they do this reversibility is that they were not able to overcome complexity but then they've made it so easy for you to back up any change and because the change is so small then it's actually easy technologically it's easy for a person it's an API okay boom something happened okay well we'll learn from it and then you can actually you know improve on it because imagine if you can do a hundred mistakes well you don't have to wait a hundred weekends so that you can do it over the weekend when all the IT instead of going with their family they have to stay over the weekend because it's that's the time of the year of the of the month when we can play within our sandbox and then I will call my wife and tell okay me and the boys we go to do real stuff and we go into a room which is war room and then we'll have all the dashboards and we can play and then we have two two days or you know like 48 hours window where we can actually you know mess stuff up learn from it and usually like we have a very small window because the backup and the update of the production takes 24 hours so it's really you know the window for you to make the changes very small so keep calm and push ups as a service there's a couple of things that you can do basically you can hire someone you know who has done it who has successfully delivered and then he can do it for you this is our methodology that we do with pivotal labs and basically the the first step we do with delivering a product is we set up a CI CD pipeline another use case would be that you can you start learning you read books like one the Phoenix project which is by Jean Jean Kim and he describes these three ways how you can iterate very fast on your products so basically you do this content continual experimentation and learning and this is not something new this is something that Albert Einstein's many of the professors and innovators they've done it in the past century were like you know the only reason the Thomas Edison came up with so many innovations it wasn't because he was the brightest because he could do more experiments per day than any other physician any other innovator so he he would just practice that and keep innovating and it's fine if he he will fail with nine out of ten if he does ten experiments that that means that one will be successful well all the others they would do it of multiple weeks multiple months because they'll wait for papers for the right process for the light alignment of the stars the second is basically the feedback the only way for you to actually know if it's successful it's not if your manager tells you okay good job if it's the market if it's the client if someone who is using your application is actually telling you yes I like it and you can actually see it in the when you push an application in the up store you know you'll have reviews bad or good but at least you'll have feedback it just happens that for many applications that are pushed into the up store on the mobile space after the first day after the marketing budget ended it's gone in the oblivion because nobody's looking at it and the last is the flow and this is by a sport coach with a very hard name where he basically talks about this flow state where you have innovators which are your engineers and the way they can get into the flow is the same like some some days of the week you know where you don't get any interruption and you just you know literally are working on something like you know for two hours and then at the end of the two hours you see a result and then whoa you know this made my day this is something that you know I accomplished I would have accomplished for a week and then there are other days where like you have constant interruptions you have constant meetings and even though you are efficient you have calendars buzz assistants like the end result well that was a wasted day nothing nothing came out of you know that that date that will influence your company and this flow state is basically where you get to work uninterrupted on something very difficult because the time and the context that you need to get in order for you to develop something meaningful it's you know it's eight hours so you literally don't need to be interrupted by calls by what's up messages alerts emails manager coming with for you pierces and that's the flow state now how do you how can you enable the fast feedback well you know we adopt agile and then we do it every two weeks but it's not it's not only that case because it's actually you have to release less change more often this is the main attitude that you should take with your product it's not that you do it every two week and you still deliver the same mess it's actually you know you start delivering small incremental changes and they will have a low risk and you can do it more often and more often so from 12 months to a quarter to a month to a week to a day the myth in Amazon or at least it they go by is that not only on the first day you actually commit something to production but you you have 13 releases per second so it constantly changes and if you think about it is like well that could be impossible but if you think about all the products that they have in their catalog all the thing all the services that they do there will be someone in a region on a product they'll want to make a change and they do it because it's possible it's just if the person wants to do that change and make the improvement so continuous integration and delivery right so I think we got it this is what we need to do okay every every granted organization begin speaking you know having the flag continuous integration delivery continuous integration delivery and then you start having you know agilist the the only way that you can do it is we all go agile right everything you know from top to bottom you know it's raises the banner of being agile if you're not agile when you're not with us so it's a nice nice image or it's DevOps everywhere it's like okay it's DevOps if you're not DevOps it's not gonna work you're not gonna do continuous delivery or you know we're gonna be out of business everywhere like everywhere is DevOps we all do DevOps starting from the CFO even CFO has a has an access to production right so you can push changes to production or you go it's all about design it needs to be this you know nice design where we go into the ivory tower and we finalize our design and everything is OOP everything is object oriented has all the principle all the practices everything that will push into production needs to be checked by our chief architect and this is my best it's actually taking this is by model IT we need to find the right ratio between our old stuff and new stuff so that we can deliver very fast and you notice the green guy so honestly you know this this sort of things were like there is a magic ratio where like you know we have two parts of our infrastructure landscape and they all have to be aligned it never works maybe on a paper maybe on a gardener analyst report you know again if you've done enough research if you've got it and asked all the right CIOs then the percentage will be right and we'll move and align to that strategy so I think it's not one it's actually three four even more things that are come together and it's very important to understand that it's actually only if you combine you know these practices these people the product then the sum of the others will be much larger than you know automatically this is the same where they teach you that one plus one doesn't equal to but it's actually 10 because if everything works together it's if everything it's in that flow then you can actually iterate faster it iterate much faster for your company you get to join the the unicorn club and let me just read through so you have to work on a product but that product is not your average project you don't get to assign project resources you do it but it's very much you iterate so the only way for you to to build something meaningful if you iterate on it so that product is continuously improved you know you get feedback from your clients you get feedback from everybody in the organization as long as you work on a product then that means that you can actually deliver value of course process but process is not you know agile come ban or any other it's a process that works for your organization it's adapted to the knowledge that is in your organization you cannot send 10,000 people to an agile or scrum certification after two days they come all certified and you know nothing happens because they well they learned how it works in other organization and they are back into your organization and it's like okay well I need this and you find all the excuses that you see in the wall so you need to adapt the process and basically improve the process and of course it's about people and I'm not talking about people that you know have been born you know like they do in many other cultures you have to be born have the astrological sign be in you know have this sign and then if your project manager also by the stars is in that sign it's like a Leo and then it's born in the year of the dragon then you will work together it's it's basically people who understand that they can they can make a change and if you look at all the Unicorn clubs they don't get new people new clone people right you don't see Zuckerberg and then clones of Zuckerberg doing all the other startups it's actually people from the same industry like former your former peers that have developed this learning process where they constantly improve and they contest constantly adopt new practices and that what keeps them ahead it's you know meetups conferences blogs hacker news whatever whatever that keeps you updated because you need to be in that constant learning so let me give you a couple of definitions what do I mean by continuous integration it's basically having a trunk and merging the dev to the trunk so you only always have you know one one trunk based one single truth in your code base so you don't have multiple branches people you know with one part of the truth other people with you know another small branch of the truth it's like one truth and then you can basically iterate on it because if you look at the way your clients use they use one single product so then it makes sense that you also maintain one single truth about it then you basically have build service that automates this stuff so it's not about having green KPIs how many percentage of specific tests it's really you know keeping the obvious for the whole group like if we are working on a new feature on a new product here it is this is the source code git clone and you have it if you want to make changes this is where you do it just make sure you leave it green after you've done your changes continues delivering this is basically once you know that that source code is at the specific location and that is the truth for everyone then you start doing a staged approach where you move it to dev where there is you know a sandbox you know you don't have big servers you don't have a integration or latency test and then you move to let's say uat where you basically integrate with the larger picture of the whole organization and then you probably will find bugs errors things that were not thought before that were not present in your development environment and then you'll move to a pre-prod environment where it's very realistic to what will happen to production where you have you know data it's not mock-up data but it's data that is very realistic to your production and things that you you will not be able to predict on the first stage because your production environment could be running for a year 20 years you could have legacy products that you know not many have a context and this is very much important about your organization that I'll go back to the complexity you you may start small your product may have microservices but it will outgrow it will happen that you know you have maybe three people who really understand how everything suits together so the story is once I joined the big bank I started asking myself okay so what's the process how do I do this dot and soon enough they they send me to all these different people sent tons of emails get to chat with IT DBAs release managers change managers portfolio managers you know because I wanted to make a simple release and then at the end of my journey for about like three months well I said okay now I get it I understand what our product does in the big or the larger scheme of things and then people look at me oh really okay so you we only had three people in the whole organization who knew about the product and the whole life cycle so we'll now add you to the fourth so you'll be the fourth person and you know to me that was like an enlightenment okay so now I get to work on real stuff instead I was put on a big mailing list where I received even more maids for people asking you know questions from what happens with the logging what happens with the monitoring like stuff that I couldn't actually bring any value because I was connected to the firehose so you'd have all the people asking so many questions and even for me you know that that counter on the mailbox keep incrementing so from you know the first week a thousand next week 2000 and you know I you know I was doing my best actually high and you know help people to understand but if the people with whom you work don't take that learning approach to you know start being curious and learn for themselves and they stick to their silo approach where like where I'm just a brick in this wall this is what they do you know if you move him to another part okay I need six months or three months of onboarding okay so continuous deployment is then basically the final of it it's where you you don't need people overseeing your build you don't need people you know you don't even actually have people who are managing your product it's basically an API for it it's something where you go push a button and it just happens and then you say okay that's good so continuous deployment is you know you have one single truth it goes through that stage approach from dev to production and you know the only thing that you have is once you commit it and they probably could happen something really bad you just have a push and a rollback button that's it keep calm and push applications as a service because this is the only thing that you can actually add value on your organization everything as a service like you can look in the market and there will be a service for everything even for chairs a service right there will be an API a mobile app someone to call that will give you adjacent which will have all the details about this chair or about this table because this is how we're moving we are moving now into a new era where we have the Internet of Things where basically what will happen is I will go to my home the electricity will light up my fridge will start talking to me talking about news and then I'll open the fridge it will have the right amount of calories the salad will be there I will not have to open beers like everything will be there you know just Internet of Things because they will know so much things about myself and everything will be API controlled that will will leave almost ubiquitous world where like we have you know a big entity that stores our metadata suggestion and then you keep interacting with APIs around the world there could be I'll have my I see I still have to go to the person so I go to the custom through us I'll still have to pass that kid so okay if we move to everything as a service I'll keep it very simple let's think about the way we innovated with electric cars the model was very easy you know as a consumer I'm interested to keep my price for gas per mile as low as possible and if you remember the whole story about Ellen mask and the Tesla it was it almost went bankrupt it was very expensive at start but he still had the confidence because the oil industry and the energy industry was there for a hundred years and there was no more innovation and then the innovation began with you know let's make an electric car instead of putting you know as many computers on our you know hybrid car and it will have too many sense many senses over it what if we could build a car that at the center of it has a computer and then we put the wheels on it so then everything is API driven and soon enough of course the prices went up of course all that the industry you know realized that electric car is the future and we will all have electric car will have even driverless car I think what you can see in the in the industry is that it's so easy to look at things in the past and make the realization that yes this is how should be but if you remember like a few years back it almost went bankrupt this is what happens in our daily life it's not like it's a flow it's a continuous flow it's not that you know from day like anodominy you know this is changed and this is like the right point in time where the whole industry from 99% who are using oil now changed to electric there's of there's of course a point of time where you know there's enough crazy people who believe that this change is the right way let me give you another example about the way we do it software with an example about car as a service so on the left we are very you know it's very easy for us to to relate it's basically I own a car now that you are all in Singapore you know that this is not a reality this is not how people in Singapore move around so then you start thinking about okay what if we can lease the infrastructure and make this infrastructure as a service so then we have this managed by the client and managed by the service provider where we manage the servicing the renewables insurance road tax garage fuel road tax driver and we basically lease it for 10 years and we participate in a bidding where like we get to use a car for 10 years well that's one way what about the next way we're like we actually start thinking about more of a platform where we instead of worrying about you know servicing renewables about the technicalities of it we are not all mechanical engineers we don't need to know everything about the car and go you know once once a Friday put a bolt on it or you know change something in the in the engine because this is the reality we we have our day job and we focus on what we deliver at best and we specialize in a specific need where the market pays us you know you know a good wage that we don't want to all go back into the garage and become mechanical engineers or auto service this is platform as a service where you have the only need that you do is basically put fuel road tolls and be the driver right have a license now the ultimate goal is basically my Uber up where I don't need to worry where I parked my car I don't need to worry about fooling I don't even know where's the net you know the closest petrol station call point A to B this is software as a service now why am I bringing all of this in terms of I work for a software company I work for a software company many years ago Adobe and then I joined a financial services company because I thought you know I'll be able to make a change because I learned so much and you know they will understand the technology is all about it so I'll go into a bank and make it a new bank because it's just technology soon enough I found out that it's actually the people people don't want to change there is a good jokes where you know you have this recurrent question like how many people do you need to change a light bulb well the now we feel in the industry the recurrent joke is that it's enough to have one person but the light bulb needs to want to change so he needs to be you know to believe in that and he needs to to know and want to change so we at Pivotal we have we go by a mantra we don't just transform how you build software we believe that software is now so ubiquitous that you can only iterate you don't only build once and then you know that will be for decades you actually keep iterating so then it's all about those three pillars of continuous delivery people process product and this is like a good example of how we actually iterate on software in any in any organization we begin with a light bulb an idea we start writing something you know could be a you know a story could be a note could be something that you record a message on your you know on your assistant and then pass it to the developers soon enough you start building software but the only way for you to know when you're actually finished if you also build tests right like what am I building what is the end goal when do I know that you know what I'm building it's finished it's done then once you do the whole testing you know you deploy it right fine enough and then once you deploy it you have people so you got a feedback you might not be good it's you know people keep forgetting that it's not about the technology where you know I can deploy a hundred times per second or you know every week it's still about the feedback he are you building the right stuff because we build organization where like if you look at our organization is we have dev we have DBAs we have testers and they are all building they go every day and they build stuff eight hours a day 40 hours a week and they do it as fast as they can and they go to the certify certification so that they build even faster the broken part is that they end goal or the feedback they never see it so they don't know if they build the right stuff or not even if you know at their position it's not that they want to fail the company or you know waste the company money or go bankrupt if you go to any you know to any engineer or employee of your company he doesn't wake up in the morning and say okay I'm gonna I'm gonna do something bad in my company I'm gonna work for eight hours you know just to break it to set everything on fire now it's like you know you go there you work even more especially in Singapore you work more than eight hours for sure now one of the reason that cloud gov is using pivotal software is that we are big believers in open source and actually we you know set a tool in the open source space where you can actually get and use it because we believe that you know any company that builds products if it's not building it in the open source community then it just it's lying to yourself because the only way for you to gather real feedback not from you know analysts where you have a maybe a monthly call you know about the yearnings it's really if it's there in the open source you have a community because community will bring diversity diversity will bring new ideas it will bring feedback this is pivotal tracker we think that you know it's a good Kanban board where you can actually write your stories and iterate on it and you have this visit great visual this is spring framework the namesake of this user group we believe that it's a great choice for you to build microservices this is concourse a great CI tool which you know the pipeline I showed you in the beginning where it has everything visual for you what's failing what's starting what's spending and this is cloud foundry the only way you should deploy software my in my humble opinion and I'm biased I'm biased but I challenge you find a better tool change your organization if you do it even better than will will live in a better world right and got a much feedback from anywhere anytime keep calm and push application as a service here's my source code run it on the cloud for me I do not care how this is a haiku by oncifakuri the vice president of engineering for Cloud Foundry and spring herring questions comments one question the analogy on software as a service and internet of things I think that is quite interesting to see them because first of all with your analogy I actually want to drive the car I don't want to be driven around so how do you in the internet of things I think this is then becoming more and more I think as long as you have an API that can you can always opt out if you don't like it and use it you know in the same as you have the policy with how many of you like to receive calls from marketers that's why the government of Singapore established do not call policy if you like by any chance to be called on a daily basis every hour then you don't need to put your phone in the do not call registry but if you want to enjoy you know you so I think it's all about opt-in and out as long as you have an API you're given a choice if you don't have it there's no choice this situation where sometimes we will be teams but we find out that product manager is not very sure what he wants so what happens is a lot of times we're trying things we are HR we are trying all these things but then because he's not sure what he wants and then kind of I think it's very like I have two kids and you know when they are small they don't run they crawl they actually fall a lot and then you know my my take on it was like I'll build cushions everywhere in my house and everywhere like it'll be like it's constant you cannot babysit a project and it's fine if you fail it's fine if you you know deliver something that it's not 100% the best that you can do it it's fine to fail it just learn from that exercise don't repeat it I think that's where cut out that's where the reversibility principle is as long as you have you know a fail safe within your production you know it's fine if your version 1.1 is worse than your version 1 you can always go back to that version right so you'll you'll do it you'll learn hopefully you'll learn something for it and then you'll improve on version 1.2 I'm wondering how you can integrate for example A.B. testing to your CIC pipeline yeah so actually Cloud Foundry has A.B. testing in built into it where like it it becomes really easy for you to segment the traffic and you have concepts like you know feature toggle that makes it easier for you to have multiple versions within production this is how the big guys are doing it if you look at Facebook Twitter or you know that they segment the traffic is the way they got the feedback is not that you know will pick you know Monday for users in Europe will pick Tuesday for users in they actually do it 24 by 7 they say okay so if your IP is from this address you'll be given this experience you'll be given the other experience it's actually for you to segment the traffic as long as you make it an API driven where you know you have a single source of truth which is your software but then you allow feature toggles to give you know a certain experience if the conditionals are right don't go and build different products I think it's one of the worst mistakes that you can do is actually integrate configuration and code in the same repository and then have multiple versions where like you know for a specific use case go to software A for another experience go to software B C because then you don't have a product then you don't have a single source of truth then you probably will fix version one but you'll not fix the others and then soon enough you'll go into that round and cycle where you keep fixing things and you don't know what's happening you don't know the you know what if you're really doing the proper thing so you need to say it's everything is job based but you just provide hopes for making yeah feature toggles traffic segmentation it's it's not necessarily something that you know it's by history like so I mentioned that I work for Adobe so now Adobe I work for a company for Omniture business unit and one of the products that we're building were the marketing campaign tools so back in the year of 2008 we're able to segment traffic for different customers and then what you would be given is an endpoint you know you are software as a service and then you'd be inserting a JavaScript and then based on your profile you define targeting you define different experience and then you'll try target you know an audience and then deliver one experience and then you'll just be able to control via your target us and I think we build a good product because up until this day Adobe marketing cloud is the leader so this is changes and Facebook is delivering these small very small increments I think this is something that they only can do because they have such a large use space and honestly I think if something goes wrong it's so small that nobody will realize it but they still have enough traffic to actually significantly understand whether a change has gone wrong if you have a small use space you probably can't do that and it also depends on the use space that if you could use something that would upset the client very much you would not be able to use these kind of patterns right? You know it's like probably in my first week or month in the job the most often answer that I would get from IT no just don't do it it's impossible because A, B and various excuses but if you look through the you know even how we release products in financial services you still have that help desk call and then you call that number and then if something went wrong with your transaction then they will be able to login in their mainframe system and then reverse it back so there's no such thing as we cannot fix it is the way that we think about it is that there needs to be a person at the end of the call you know there needs to be a person who will be overseeing these emails and this distribution group or you know but the problem is it's very hard to scale people you know at any time I'll be able to reply to one person and what happens if I have you know a hundred users a thousand users and this is my single point of failure the moment I have an API then it's just a matter for me of course it could be wrong you know it will not be successful from day zero but like you said in the iteration 88 I'll probably hit the jackpot and this will be the API that everyone will use it the same way we thought about internet and online banking how many people were like using online banking when it appeared and said like nobody's gonna do any transaction on web which is ridiculous or even the code from IBM where like there will be five machines worldwide that will be used like nobody else can need or needs to use a computer there will be five machines worldwide or there's enough memory you know this 512 bytes that everyone needs for RAM by Bill Gates so keep iterating Cata and the court and I think for all intents and purposes and I give this advice to a lot of people who come and ask like on what should I focus what should I learn is you know find something really challenging find something really difficult to do and get it done because you'll spend the same amount of time and hours doing something unmeaningful reinventing the wheel instead you are giving this time frame that you better use on something that is meaningful for you for your company for your family stop wasting your time where it's like I'm gonna go and reinvent the wheel like how many people go and reinvent the wheel on their daily basis How do you advise like to create this culture for you when your teams teach for good how do you create this culture for you because some people don't really want to improve how do you want to do that you know to learn about Java Spring Concourse but they'll definitely find something in the STEM research Biomed Tech even like know how to make a better toy for their kids and as long as they find something really interesting you know meaningful like you can always take from people you cannot give and you know they need to want something really so that they can actually make it as part of they need to own it like this is one of the reasons where like you send people for certifications you get zero value and you only get value from those people who really believe it and they wanted to sign up for that learning session for that certification Thank you, thank you all Last slide So this is RAN.Pivotalite I mean the best way you would go to this slide and then you have the link for the presentation hello-cd.cfups.io and then if you are technologist and you want to make a change within your product then there is this great workshop done by my colleague Carlos who left migrating a monolith application to cloud native but for the interest is this is how you, this is a great session by Onsi Fakuri where he talks about how you iterate on software this is a great session by Josh Long where he life codes an application for one hour from scratch and the end result is amazing on what you achieve in one hour and the last but not least is concourse which I think will change the way we build software Thank you