Managing Cloud Infrastructure Services in a Distributed World by Claudia Galvan
Video Transcription
Welcome everybody. Thank you so much for coming to, to my talk. Uh I'm Claudia Galvan and I'm your host today. Uh Just uh very quickly everybody remembers uh 2020.It was like the worst of times and it was the best of times, just like in a Dickens model novel. And uh basically, you know, 2020 we already know about all the bad stuff, but we also know about all the good stuff, right? That happened. It was this tipping point that basically launch a huge hyper uh technology uh boom. Then, you know, people working remotely uh uh distance learning telemedicine and you know, everything kind of like exploding, right? Then this was a very, very important time for many of us then.
Um as part of this process, then I'm basically working at Oracle currently. Uh I'm the director of Technical program management. Uh I'm also a chief of staff. Uh I have, I have been leading uh the cloud infrastructure, cloud delivery team for the last year and uh you know, I have a team of uh that is distributed globally. It's uh all over the US Canada, the UK uh in India and uh before that I have been in software development, working with international teams all over the world and uh at different companies uh at Adobe Microsoft and believe it or not many, many years ago, I was at Oracle. Uh This uh has uh uh that this allows me to present about this topic. And, you know, I'm just going to make some assumptions that uh many of you don't really know what cloud infrastructure is. And uh then I give a quick uh a quick uh update on that so that uh you basically get uh a very high level sense about what it is. Uh And also what is the de model because this is kind of an integral part of working in a, in, in the cloud.
And last, but not least, you know, some of the best practices that uh that we use, um not only at or A O but in the industry to working with a distributor team. Uh then let's talk about cloud infrastructure in a nutshell. Uh I like to think about c infrastructure in the analogy of cars versus roads. And basically, uh we all know how to drive cars, right? And uh we basically use the roads all the time. If somebody were going to ask us build that road, you would be like, well, I have kind of like a high level understanding of how roads work. You know, they have, uh uh you know, they have lines and pavement and curves and, and so on. But basically all these uh SAS applications are cars and all the roads is the actual cloud infrastructure. Then I like this analogy because if you think about it, uh, now we're building roads for cars that have not been invented yet, then hopefully that blows your mind. But, uh, you know, it actually blows my mind uh every, every single day, then I'm focusing on building roads as we speak. And uh just very quickly, some of you may already have seen, you know, this type of uh uh diagram where basically uh talks about the different levels of um uh involvement from a customer perspective in the cloud.
Then, you know, many times like us, you know, we use Zoom and we only see uh Zoom as an application, right? But there's like all the different uh uh components in the stack to make the the cloud work. Then in my case, uh you know, I am in between this and this space. Um And uh in order to make the cloud work, you know, you're going to need compo as uh as compute as a storage as network that we all take for granted, right? Uh There's going to be like a middleware of um uh you know, basically uh libraries and API S that, that basically enabling the, the SAS application to work then and I'm just giving an example of Zoom because Zoom works on Bronson oracle. But uh And that's a little known fact. Uh But then anyway, then that's uh that's, that, that's basically, it is uh it is a stack of services that all together allow uh SAS application to uh you know, to have AAA very seamless uh U I experience to the end user. Then let's go to the next uh to the next one. Yeah, then one of the big things about the cloud and we take for granted is like when you start zoom, for example, you just basically assume it works. And this is called 99.999% availability, right?
It's like we, we never almost never see Google, like not getting like an immediate response or Facebook, then anything that is working on the cloud, you know, the expectation is that you're going to have to keep the lights on all the time. But how do we do this? Right? Uh Well, in the, in the good old days, you know, you had like different people performing different functions like um uh you know, you had your Q A department, you had your, your engineering department and then you had like the ops team. Well, kind of like in the more um uh more current uh development model, basically the same engineer is responsible for like everything for all phases. And basically it goes, it goes through a cycle and this is going to is called continuous improvement. And basically this is what keeps the lights on. This is uh basically what allows to have 24 hours, seven days a day, seven days a week uh availability. Then uh you know, there are systems when uh you are a program manager to basically manage this the box models, right? Number one critical thing is a clear understanding of roles and responsibilities, you know, from the engineer all the way to the uh management chain, right? From the people who are in customer support, getting the uh you know, the getting the call from the customer that something is not working and then basically triaging for the severity of that particular incident.
And uh you know, everybody needs to know like what they're responsible for what they need to do and how to do it. And you know, in order to achieve that you have things like uh room books and room books basically are step by step instructions that allows anybody from like an IC two to basically solve a problem, right? And if the IC two cannot figure it out, then you have like a clear escalation policy that is going to allow you to uh maybe uh get the attention of a subject matter expert who is going to be able to, to help out. Then in order to do that, you know, when you assign an engineer, you don't necessarily uh uh have that engineer working on features at that particular point, you know, they are on call and that's like their only responsibility for X amount of time. And um, and you know, they know what to do and if for some reason they can't manage uh and solve the issue, you know, there's a clear process about how do you hand it, how do you hand it off? Then uh the other big thing about the box is to have clear metrics, right? We talked about uh availability and fulfillment is uh is basically being able to fulfill a request.
Then when uh uh so as a service application goes and requires a specific resources, then we're going to be able to provide through those resources to that application. And then uh you know, being able to fulfill the request and then latency, that's another metric that we keep in mind that is basically to minimize the response time. Then the expectation is that, you know, when somebody clicks on something, you're going to get like an immediate response.
If you are basically waiting it to load or to grab data from the database and all that, that's not a set of acceptable, right? We have to have, you know, a very, very small uh number for, for latency. Then um then with that, those were just the basics. But then let's talk about working uh with a global team, then just, you know, to, to set the stage. Um we uh a few, a few jobs ago, you know, I actually worked for Adobe and uh I was, uh, the majority of my team was either sitting in Seattle or in India. And, you know, I traveled to Seattle every week and I didn't travel to, to India every week. But, you know, I kept up, I kept having calls and all that and I always thought that I did a good job. Right. I was in the headquarters in San Jose and, uh, I really didn't think, uh, uh, much about what, how would it feel to be uh remote right now? We are all remote and now, and you know, it's kind of like uh we all are sharing that, that experience. But when somebody is sitting in headquarters in person and then people were on the phone, you know, it was a very weird experience.
Uh When I moved to Microsoft, I was actually in a remote office in Silicon Valley because the qu the headquarters were in um in Redmond, Washington. Then I, for the first time I actually experienced what it was like to be remote. And at that point, basically, I put together like a um a task force with other people who were remote that had lots of experience. And basically, I started to compile like uh a set of best practices to working with people who were not in, in the office. Then fast food work to 2020. Now we're all remote. And basically, I have uh been evolving this uh these best practices based on those original ones that I created with the Microsoft team. And then I basically ended up uh uh just adding as I, as I gain more, more experience, then I, I called them, you know, I start with uh just I'm just going to go through them very quickly and then go in depth to each one. But, you know, basically alignment, conceptualize, conceptualize and trust frame the communications fair trade champion, uh and recognize and celebrate. Then with that, then let's talk about the power of alignment then especially now that we're all remote, right? It's so critical to ensure that we're all aligned into like the same, the same goals. Then a few years back, uh there was this organization called Random hacks of Kind Kind.
And maybe you have heard about it, they basically organized uh ma uh hackathons and these were like three day events where basically people gave away their um weekends to try to solve uh a worldwide problem. Then let me give you an example like uh earthquakes in Nepal. Uh this is not a worldwide problem, but it was a very impactful problem at that time. And how would people communicate? Then we needed to create like an emergency app and we all work, you know, day and night for three days to try to create this this solution. Then this is like we all wanted to fulfill this goal in a very short period of time and we had this alignment right. Then uh other companies uh have uh reached out uh and use this power of alignment and using the, and using crowdsourcing. For example, when Facebook just started, they basically reached out and they said, OK, we are inviting everybody to help translate Facebook. Do you remember that? And basically, we, we did that. Uh Many of us, we went, we donated all the time just to, to translate. You know, many of us have jumped into the wagon of try to, you know, make more money and say, hey, I'm going to rent one of my rooms and, you know, and then crowdsource the effort of um uh providing uh accommodations to people around the world. You know, it's a very clear mission from uh from uh air BNB. Uh And, you know, we can go on and on like, you know, just uh donating money just to allow people to uh to fulfill their projects.
Then this, this single goal, you know, allows, allows the teams where they are, allows the people to where they are to basically have um a mission and a record and, and deliverables. Then uh the second best practice that I would like to talk about is in context and in trust, then imagine now, you know, that we are in a virtual culture, right? Uh Some of uh some people are jumping into the uh the the metaverse, not me yet. But uh but basically, you know, the latest statistics is like 70% of the people are working from home now. And actually a big percentage of people are never planning to go back to the office. I think the last I heard was about 65%. That, that was crazy. Then 95% of the meetings are, are now virtual. And uh you know, we are in a virtual culture, you know, and uh using uh using, using the the different tools. Then how do we establish that trust in this environment? Then uh one thing that I try to do is to set up meetings that I call them, meet and greets, then I get started working with a team. And uh I basically uh provide, you know, my own background. I said, why I am part of this team? I tried to learn about why they are part of this team, you know, what do they want to accomplish and how do we want to accomplish? And I have this one on one meetings on a regular basis with basically a lot of the people in the team. Why?
Because, you know, you want to build this like glue and relationships and trust uh with the team. So that when you need something, you know, they're going to be able to help you. Then uh uh a while back, it was like this book that is called uh Three Cups of Tea uh that you may have heard about the Greg Morson. But he basically uh went, he, he, he was a climber that, uh, basically stayed in Afghanistan. He made, um, uh, he, he saw the needs for schools for girls. And this was before the Taliban. And he basically, at the very beginning he was like, oh, my gosh, you know, I want to build a school and everybody's like looking at him. It's like why, you know, it's all that. Then he, he, he took the time to learn about the culture. And uh basically, it was like the first time you share your tea, you know, you are a, you are a stranger. But the second time, you know, you become a guest and then the third time now you're family, then, you know, investing in building this context and trust with your team, especially now that there is no time to have coffee or lunches or things like that, then spend, spend, you know, set one on one, join the meeting just a few minutes so that you can basically say, how's the weather, how's your family, how are things going?
What do you do over the weekend? You know, those are like uh great practices. And uh then the, the next, next practice is about framing for clarity and uh framing for clarity. It's about just a framework and, you know, it's really, really crazy that we take for granted a lot of things, you know, it's like who's doing what, you know, the, the who's doing what what are the priorities, how often, what, what are we going to, to get out of uh of uh communicate of this communication?
Then uh you know, when, um if many of you have worked on uh A G teams and, you know, there's like a clear definition of who does what, right, who is a product owner, who's a school master, who's a team, then, you know, that's like no question about it. Some of the teams are more organic. If you want to men to talk about that, then providing some type of like a, just a matrix where you basically says, who's going to do what when and uh it's uh and what, you know, who's going to be participating in this con conversation is uh super important. Uh one area where uh I learned kind of the hard way was working with uh with Japan. Then I used to manage uh the all the requests coming from that where a Japanese team, our team uh that resides in Japan. And basically they had two types of issues uh crucial and critical. And it's like, OK, I'm not really sure what that means. Then we established uh uh a process where we basically said, OK, just tag your request. If this is an idea, just say this is an idea. If this is a question, just say what this is a question, right? Or if this so urgent, then just say this is urgent, then you know, you very quickly start setting up the, the the framework on how you're going to be managing your priorities, right?
And then the second thing is how often are you going to communicate? And many of us are struggling with this now. Why? Because we're all in different time zones, right? I, I have been doing this for ages now and uh you know, the time zone issue is always a pinpoint for everybody. It's either somebody has to get up early or somebody has to go to bed late. Um And uh you know, we ended up, I mentioned, you know, I have managed teams over and over again in India. Then the most important thing, especially with teams that are 12 hours ahead of us uh is to ensure that we're not overwhelming them with meetings with everyday meetings. For example, Fridays for the India team are no meeting days. We really don't have any meetings on Fridays. Why?
Because we don't want people to be working like at 9:10 p.m. on a Friday, right? Uh The second thing is that we basically provide uh start to create a cadence, what, what it makes sense then instead of having meetings every day, we have like meetings maybe twice a week, right? And we complement with uh some written communications, right? We talked about that we just a slack, then we do updates on Slack or on email and then, you know, we basically maybe just one maybe we just have like one conference call and we never go beyond like working hours if possible because it just, uh you know, you have to put yourself on the other people's um shoes.
Uh and you know, now we're all doing zoom or teams or uh you know, meet, meet, meet up everything is, is happening then more than ever it's so important to like have an agenda, ensure that we have the meetings. One thing that is really important that is happening now we start the meetings five minutes after the hour. Why? Because we're sitting in front of the computer and it's like back to back meetings and it's like you barely have time, go, go to the bathroom, have your cup of tea, you know, stretch. Then we basically start the meetings five minutes after make sure that we clarify the objectives of the meeting. And we basically, in many occasions, we record and we have like a an area where people that cannot attend for some reason. Then, you know, they can access the recording and then we summarize then which format and many of us, you know, are using and I'm just giving examples, right? Uh Outlook. It used to be like we used to get like hundreds of emails now, right? Ok. Guess now, guess now, guess what, 2022 now we get hundreds of slack messages uh a day then you know, the way that you prioritize, you know, I have my channels and then I have my starch channels.
Then I basically my star channels are the ones that required my response and that I'm monitoring. And then uh I basically check email maybe in the morning and at the end of the day, and I actually have a notification that tells me that, oh, my manager sent me an email. Then it basically on my um earphones, I get uh a notification that I got an email from my manager or any of the other important stakeholders that I need to respond that way. Then this is something that I, I have been, I have been doing and because we are in a global culture now it's so important to keep it simple. You know, English is my second language and at least once a week and I have to ask people, what do you mean with this? Because, uh, you know, we work with people from Britain and they have like a different slang and they use this slang. Um I think 30 years in the US and I still don't know the UK slang. I'm sorry that I have to go and ask them to keep it simple, you know, uh you many, many of, you know that, you know USA today, this is what the people say.
It's written for like eight graders, this is what we need. Um And then, you know, we talked about uh not using slang, right? Uh And many times because we're in different time zones and all that, we really have to provide context, right? Try to use formal speech if possible, try not to try to avoid a slang spell out of the abbreviations because you know, there's so many, there's so many and not everybody knows and use English don't mix and match and, and assume people know, do you know that some people use Latin in the communications and I'm like, have to Google this.
I'm like, please just use English and then um OK, then moving on, then let's talk about the clear outcomes. Um And uh OK, then on clear outcomes, you know, it's basically what we want to accomplish from this meeting. Then even before this meeting, I was like, I was taking notes on conference sharing my screen and basically say, you know, these are the next steps and just type in the next steps in the conference page so that it is absolutely clear what we want all to accomplish. Uh And then let's talk about fair trade. Uh This is super, super important. Uh you know, when, when before just that this is like before 2020 after 2020 remember when headquarters people used to like get all the like juiciest projects and everybody used to get like the leftovers and the crumbs. Well, now, you know, you see this is a forcing function, right? We're all remote then, but we still have to ensure that, you know, basically everybody who's working on a project, you know, gets the right visibility, the right ownership and that, you know, people at the end of the day, they feel confident that they're going to be uh able to go and say, you know, I was able to accomplish this, you know, and that I was not just working on, you know, busy work that I cannot put on my resume.
Thank you. OK. Then last, but not least it is about a champion uh nurture as a state. What does this mean? You know, we, I always say we're all like we all a village and especially, especially women and minorities, right? We, we always sometimes uh uh don't, don't spend the time on raising our hands and we're just like heads down on our projects. Then I have learned, you know, that it's so important just to ensure that you, you champ your champion the uh you know, and you recognize the, the rest of the team. Why? Because uh you know, at the end of the day, everybody, you know, just having, giving AAA small, thank you to people a small uh you know, uh recognition or writing and copying uh uh their managers about like all the hard work that they have done and that you appreciate. It just makes a big difference at the end of the day, you know, they're going to be a, they're going to want to work with you, you know, in uh like the next project right now. Yes or no. We have been heads down the last, you know, just the last two years going like crazy working virtually. And how many has taken the time to basically go and send an email and copy the manager and say, thank you for all your hard work.
This is your action item. By the way, if I have my conference page, I will go and sell your action item. Make sure you recognize all the people who are you are working with. Because at the end of the day, if everything comes back, right, people will will pay back as as well. Then today, you know, I give you a quick uh overview of cloud infrastructure. You know, if you're going to remember anything about cloud infrastructure is the cars versus roads analogy, then cloud infrastructure is the roads. You know, the other thing about the box is basically the glue and ensuring that everybody has like a process, clear roads and responsibilities, escalation paths, uh documentation so that they can solve the problems and that there is like the the people who are working on supporting the site, you know, have all the tools to maintain that 99.9% availability then and and last, but not least, you know, we talked about working on the on a remote uh on a remote team and that, you know, moving forward.
This is this is the new reality. This is the future of work. Then we basically need to switch uh how we work with our coworkers and uh and our extended teams and basically ensure that we provide always the context that we have a clear framework on how we're going to communicate. And that we don't forget to, you know, to basically ensure that everybody gets a right piece of the pie and that we recognize all the efforts. Then with that, I'm going to, oh yeah. Before I forget, make sure you connect with me on, on linkedin. Uh It's always good. It's a great, great process. And uh I'm opening these for questions now. If you can uh ask uh ask me questions on uh on the chat, please. Thank you. All right, then no questions. Then I want everybody to say, I, I really want to appreciate you for coming to my session. Stay in contact, connect with me on linkedin and good luck working with uh with your distributed team. Thank you so much. Take care.