 Kalyan, let's start with a little bit about yourself. Tell us a little bit about who you are, what you do, and where are you currently working as. So, I'm Kalyan. I work as Lead DevOps Engineer at Media.net. Media.net is an ad tech company which is operating from Mumbai, Bangalore and Delhi. So, I'm basically from Chennai. I graduated from Gindi Engineering College. Then after campus I joined Media.net as a DevOps Engineer. And yeah, it's like five years at Media.net and we're still exploring and we still learn new things. I'm sure a lot of people already know what Media.net is into, but for the benefit of the others who do not, can you tell us a little bit about what Media.net does? Media.net is an ad company which does contextual advertisement and domain advertising. We are like a 2,500 member team scaling rapidly and recently we got through an acquisition for close to 900 million dollars. And our primary customer base is from US. We serve US and Canada, ad tech US and Canada. And right now the team is looking at real time bidding and head up bidder and other horizontals in the ad industry. Right. So, you mentioned that you work with Media.net as a Lead DevOps Engineer. Can you tell us a little bit about that journey of yours into becoming a DevOps Engineer? So, what should people just out of college who wants to get into this field, how would you recommend they do that? So, I think a DevOps Engineer is a kind of a field where you could excel if you're actually interested in a bunch of topics like the OS stack, networking and kernel stuff. So, these are some things which during campus placements and stuff people don't have, people usually ignore because they are pretty sure like you know a proper, since you asked with respect to college I'm talking about this. Like if you are interested in coding and solve algorithmic puzzles, it's kind of people ignore this part. Algorithmic efficiency is really important. Algorithmic efficiency is the only way you can optimize the cost. A brute force will take costs out. Similarly, there are a lot of things you can do on the kernel level. You can do on the network stack which can optimize the whole flow. So, there are a lot of things we take for granted. Like it could be as simple as spawning instance in AWS or opening even Google.com. There are too many things that happen behind it. So, if a person is curious and he actually reads about them, what happens when they open Google.com? What is DNS? What is a service they run for DNS? What is the web server that they are using and how do they run all this? And who cares about uptime? Then you are a right person to be a DevOps. You have all the skill sets to be a DevOps and you can give it a shot. So, that's interesting. Can you tell us some of the challenges that come with this role? I mean, a lot of people want to get into this industry. So, obviously, they should be aware of what are these calls at 1am at night saying that everything is down and you have to fix this. What are some of those challenges that DevOps engineers face? So, even this conference, there are a lot of philosophies that are coming up. So, this is a new field where there is a transition between sysadmin to DevOps, RSRE. So, everybody is not clear which is the right part to go ahead. So, once you join the field, you will be one of those persons who are trying to bring in some of the philosophies which might eventually be adopted by other people as well. So, the 1am calls do happen, the on-call rotations do happen. But as you grow, I think people stick to best practices. So, the best practices make sure things don't break for silly or simple mistakes. So, mistakes can be two ways, one is unexpected. So, if there are expected mistakes that happen repeatedly, it causes you fatigue and repeated calls will just bring your energy down. But if it's an unexpected mistake, at one point it also gives you happiness that you solved something today. It happens as though you are like a detective, you solved something behind and you solved it. So, that sense of satisfaction actually compensates the 1am call. So, it's just the number of calls you receive at 1am matters, it's not the calls that you receive at 1am matters. Yeah, I think that makes sense. I mean, best practices obviously reduces a lot of unexpected 1am calls. And I sleep after 1am, so I'm fine. Can you tell us a little bit about your talk at this conference? So, this conference I presented about, so it's a cloud server conference. So, by the time I came I know the people here will be pretty much working with the public cloud and most of the talks will be related to public cloud. But there's some subsequent amount of investment we did to run our own private cloud and there are certain design philosophies we took, which I thought could be used by people in public cloud also. And some of the text that we used in private cloud lets people to understand what the public cloud people do behind the scenes. And we just take it for granted, but there are a lot of difficulties in achieving them as well. So, this talk shows how we try to achieve some of those fundas that the public cloud is actually giving it. So, I thought this would be one of the platform where I could actually share the philosophies, the tech we did it. And it's a kind of moral boost for the team as well. So, we are giving it to a bigger audiences, whatever we have achieved. And by the time we sit and write something like this, we also get to know what we thought initially as a goal, where are we right now. And so, it's a kind of like once you document something, you actually get a clear picture of what so far has happened. So, I think my last question would be about why do you think speaking at conferences for, you know, at developer conferences for this domain is important? So, like I said, personally for me and my team, it's important because we get to know, it's kind of a review for us, ourselves. We just sit and we talk and we come to know that it's a kind of a self-review for us. We came to know how far we are from the goal. That's one thing. And second thing, conferences like this, you have certain philosophies and you have worked for you. You're sharing with people. So, most of the devops stuff is kind of open source and it's built over it. So, even your philosophies can be open sourcing a philosophy today, which again can be taken by somebody. So, it's limited by your scale and stuff. Another team can take it and modify it for their scale and restructure it and use it. And there are some problems which even I have a feeling that can't be solved technically. And when I came here, people have, I have seen like people adopted certain funders to make it work as well or it not 100% work but some benefits out of it. So, that gives a path for something which we initially thought it's not possible to do. So, it's always exchange of ideas helps you change your opinion on something. Sometimes when you are just alone, you are opinionated, you get more opinions, you open up and I think that's the main reason why there should always, it need not be a conference but there should always be a platform where like-minded people share their ideas and develop over it. Absolutely. So, the day isn't over yet but my final question is what has been your key takeaway so far in this conference? So, I have some takeaway from each of the talks. I talk to those speakers as well. So, I like the part so the last talk had about lessons at index operations. So, they talked about the operatability review because every team has a fragmentation issue that comes in because teams becoming decentralized and with agility needed. So, this operatability review or something of this sort actually helps you the same thing. You sit back, you think it's like an approval process. So, you think something, you fix it and it goes live. So, again you reduce silly mistakes, you reduce known bugs, you are not reinventing the wheel and everything. So, that is one thing. I fresh disk co-founder stock. There are a lot of the takeaways there also like one of the thing was he talked about blue green deployment. So, where he said how the DB servers are not part of blue green deployment. So, I always felt people little higher in the management usually are not hands on. At least I have seen like that but here it's some person going so much intricacy is something which I learn at least like in future. If you grow also you should be hands on to actually motivate the team to work that gets recorded.