Episode Transcript
[00:00:00] Speaker A: Integrating finops in the technical side and in the software development lifecycle is one of my favorite topics to talk about. And today I bring two experts on the field to help me envision how we can make this better for FinOps. So we have Ernesto, the CEO and co founder of Glacity. Ernesto, how are you doing?
[00:00:18] Speaker B: All good, all good. Thank you Victor for having me here.
[00:00:21] Speaker A: Thanks for having me here. It's always a pleasure and honor to have a Spanish representative in the FinOps ecosystem and I'm super glad to have you here today. And we also have Dustin who's doing finops in kpmg. How are you Dustin?
[00:00:35] Speaker C: Doing well, thanks. Happy to be here.
[00:00:37] Speaker A: Pleasure. Pleasure to talk today and you know, I'm going to start directly with probably a hard question that I would like you to answer. So Ernesto, in your opinion, what is the hardest challenge that is to be addressed in finops these days?
[00:00:55] Speaker B: Well, for me the hardest is always the human component. I think that addressing that part is like really difficult because there is like a lot of intangibles on that part and rather than the technical thing that more or less we can address with tool or with different mechanisms, I think that the human component is like the most important or the most difficult. Aligning interest and all that.
[00:01:20] Speaker A: Yeah, we have the irrational component. We don't work like always by the rules and we do whatever, sometimes whatever we want. So it's difficult to get some buy in. Right.
[00:01:31] Speaker B: This is the reason. It's like sometimes there is fears or aligning, no expectations or what do you want to get? Not all the people that participate in the FinOp process have the same interest and this something needs to be aligned. It's not the same like engineering or finance or leadership.
All of them need to converge.
And this is one of the biggest challenges for me.
[00:01:56] Speaker A: Definitely, definitely. I think this has been formed from the very beginning of the practice and it will always be from going forward and that's thing. So what is the hardest challenge on your side? Are you different? Are you agreeing with Ernesto? Do you have a different approach to the practice?
[00:02:13] Speaker C: Yeah, I have a answer that converges with Ernesto's and it's interesting. We did not plan this, so I lead the Cloud FinOps team within our audit line of business at KPMG. And so I should say up front that we're doing internal finops. We also collaborate with our colleagues who serve clients. But. And then the opinions I share here are my own, not those of kpmg. But coming back to the question, I think it's specifically human. It is finding out what motivates people.
And we're trying to work with stakeholders, most of whom are not directly accountable to us. We're trying to influence them, we don't have formal authority over them.
So what is it that motivates them? And then if you take that another level up to leadership and you think about motivation, what are their goals for the cloud? And sometimes there's not a clear answer on even why are we in the cloud? And so you have to do a lot of research and run experiments to find out what motivates stakeholders.
[00:03:14] Speaker A: Definitely. I think you, you get into an interesting point about this. The motivation, the incentives is something that, especially related to engineering, I've been hearing like a lot that incentives and motivation is one of the drivers of the finance practice.
[00:03:30] Speaker C: Right.
[00:03:30] Speaker A: So because, you know, we have seen like there is a lot of optimization providers and a lot of optimization, like you see all the optimization that you can do in your AWS dashboard, in your GCP dashboard, you can say, yeah, you can save, you are overpaying over here and you are overpaying over there. But you know, Ernesto, maybe from your experience in developing the tool and the practice, the problem is not the abundance of recommendation or identifying the recommendation is to actually prioritizing them. So how do you think this prioritization and this management of the backlog of optimization can be solved in the future.
[00:04:09] Speaker B: Of finops who solve the problem of the backlog? I think it's very important to have a systematic, systemic approach, systematic approach to start to collect them.
And I think it's very important also divide in different phases and ensure that we take the things that we can do and move through these stages later.
And most important as well, is able to reduce the noise, ensure that when you fill the backlog, you don't have repeated items on that. Because when you want to make the decision about what we are going to tackle next, we want to do an assessment and ensure that we are taking the most profitable or the thing with less risk or the things with less effort.
Ensuring that the decision that you are making from the leadership of FinOps is aligned also with the capabilities that your teams have later.
I think this is an important point to tackle the backlog.
Definitely.
[00:05:15] Speaker A: I think that especially in the engineering side like we have, if you come from, from a devil's perspective like mine, you have a lot of noise sometimes with the alerting, sometimes you get a lot of inputs, especially if you are monitoring stuff. So you, you are familiar with the concept of alert fatigue. So you have like a lot of inputs and then you don't know like what's relevant, what's not anymore. And, and you know, making like a prioritization, a proper prioritization of that. Maybe applying, you know, automatic tooling or maybe applying like a pm, let's say a PM for finops, like, okay, let's prioritize this and that over, over there in the backlog can definitely, can definitely help a lot. So, and Dustin, maybe moving on on that is like, what have you seen like in your team or in any other teams in that you have worked with getting stacked at, in processing this backlog in what is, you know, preventing this finops flow to, to work better in these days?
[00:06:13] Speaker C: What's preventing it? It would be some of the things Ernesto talked about. It is noise and it's, it is patterns of communication that are not aligned with how people work and how people think.
You know, I think people can handle a certain number of inputs and then beyond that they check out. I also think if they hear messages, they might hear the same message from multiple people. That's not helpful. They need to know who it is they're supposed to be listening to.
And so I think some of the things Ernesto talked about are pretty interesting. He's talked about the noise and I think cutting through the noise often you follow the money. Whether it's a reactive optimization, that's really high value or financial risk because right now that's a small workload, but tomorrow it's, if it grows exponentially like an AI workload, it's going to be very risky.
And then I think understanding the business cycle. So there are times when a development team would ignore your recommendation because of an upcoming busy cycle, whether it's Black Friday or busy season in our professional services industry.
And so I think understanding what type of communication and what frequency and what time is critical to make sure that your message is heard and acted on. Not just heard, but acted on definitely.
[00:07:40] Speaker A: I think action like, you know, I think at the beginning you can get informed and you can visualize, but you need two people to take that action and to take it like as an associate, systematically and frequently so that it can be applied. And you know, you brought a very interesting point which is like the integration of, of all of these and solving it early and not making it, let's say reactive, a more proactive approach. So maybe Ernesto, you have an idea about how finops can be integrated into the software development lifecycle, how it should work. We have been hearing this initiative about tflat but in your perspective how these finops can be integrated into the lifecycle of any business that is doing software.
[00:08:28] Speaker B: For me here, I think it's important differentiate what is the part of the engineering for FinOps and of FinOps and what is the rest. No, I think this is not a process that starts in engineering or should start with engineering. I agree with what Dustin said before and I think it's one of the steps that people get stuck really is like isn't it in the assessment, in knowing assessment and sponsor the process, the people that is receiving these inputs.
Usually most of the times of course it's engineering but they are not the sponsor of the process. They are the targets of this process. If you talk in the SATA point of view and I think that the best way to integrate that is allowing engineers being working in the place that they are. I'm engineer myself as well. I've been in infra engineering for a lot of years. And when you have your system built this is like a rainforest. You don't want the people like put the fingers on it. You have all these decisions about what tools are inside, which one's not and what is the process because you are iterating all the time on that. And I think that the best integration comes when there is a team like moving the process and then these actions are arriving inside the normal life cycle of the software development.
Like other things are coming like security items or, or some other items not that the, the business can, can bring to us. I think this is the best way to do it.
[00:10:01] Speaker A: Yeah, I think, I think you, you, you raise a good point. Like the system, you know, building like everything is integrated. You know as, as you mentioned we, whenever we have like all the tools in place we want to have everything like you know, control and we use our process and we use our data. It makes it super easy to, to handle.
Definitely. And you know, makes it.
If you have, you know your slack, you know, your CI CD tool of stack, you get better and better using these tools and you get more efficient with using these tools. And definitely I think finops should definitely be integrated in the cycle like either with a tool or apart from tools like Classity or others, but also from processes like okay, here is the responsible. So for example product that is a person that should have some responsibility when you are designing the process, when you are prioritizing the backlog that we just mentioned. This is a pro thing and a project management more of a thing than an engineering side. What's the prior should we prior or not it's the management. Is the leadership about okay, is this a prior like saving some money or making the applications more efficient rather than having more features that are going to.
Definitely. So the prioritization is definitely a good thing.
And continuing on with that, what you highlighted, maybe Dustin, you can tell me a bit about what have you seen that are the practices, the guardrails, the automations that within these systems of the development lifecycle work best for you, your team, other people that you have seen? What are the tricks and tips that, that you have over there?
[00:11:52] Speaker C: Yeah, so I will speak some from experience on this and some from theory and I would say that our team is pretty early on in this, I think also it seems like the industry is pretty early on.
So I contributed to the unit economics white paper a few years ago that the industry developed and there hasn't been, it doesn't seem a lot of momentum in that area in the industry it seems similarly designing for cost, cost aware design.
Actually somebody on our team is the leader of that working group within the FinOps Foundation.
There's been a little progress there on the white paper and academic content and I think within practice there's also been a little progress. But we have a long ways to go.
And I want to connect something Ernesto said earlier with one of your modules, Victor, from a, from a summer session that you did, the one that Stacklet sponsored.
So Ernesto used the word repetitive or patterns. And so if you think about reactive optimization, if we start doing the same thing multiple times.
Well, first of all, if you start doing it multiple times, then hopefully you could have operations start to deal with it and let them own it and not have finops bring it to them. But if you start doing it on a repetitive basis, that's a signal that you probably can design it right from day one.
And so there was this idea, I think I want to say Matteo in June, Victor on the event or maybe Van talked about this, a feedback loop and how if the left side of the slide is build and the right side is run, if you start observing repeated inefficiencies on the right side, that's a feedback loop of how you ought to be provisioning.
And so what we've started doing is having cost based parameters reflected in our terraform templates. And so that takes away the requirement of engineering even to know what good looks like because it's done by default.
That's one thing.
There's also a couple of other tactics that are, I think debatable.
So infra cost exposes cost impact before engineers make the change.
And their value proposition is, of course, that regulates behavior. Some people will say the engineers don't care about cost. I think that's a cynical point of view. I don't think that's true. But there's, you know, a spectrum of opinions there.
And then, of course, you have governance platforms that can actually detect violations of policy and let people know or fix it after the fact. So I. That was a broad answer there. But there's several aspects to this. But I think to me, the theme is that we're very early on in terms of achieving what's possible.
So I think certain tools in the market, including Ernesto's, talk about detecting deficiencies in your code and using AI to promote and propose code fixes that you could just deploy. That's a really fascinating capability. I think we're on day one of day 100 of that. So there's so much more we can do here. And I think it's going to take a good collaboration to progress to what's possible.
[00:15:07] Speaker A: I think that what you meant to you made. I think an inflection point on this is like, maybe I can dive deep with this with Ernesto is like, you come from the tooling side. You have experience on developing the Glass City tool. So what do you think is the superpower or the Hans capability that Dustin mentioned that can be achieved in the future with tools like yours?
[00:15:33] Speaker B: This is something that we are working on and that we have partially developed, and it's super important.
First is in the initial phases, like create this awareness. This is cool. Other vendors are doing it, but at the end, this is something reactive. Of course, it's always better, far better. Fix the problem before the things are deployed. Because at the moment that something is deployed, then we end up in a nightmare. But the last step of these five steps that you need to follow for always having optimized these feedback loops. Feed the feedback loops and keep the ball rolling all the time. This is super important because at the end, it's the same that we do in DevOps. For example, when you do a postmortem, you want to understand why we have been working tonight on fixing this problem. You want to do the same finops. You want to see why we are always cleaning. Well, because someone is coming dirty every day, then, well, let's stop him in the entrance.
It will be the same if we learn that we are deploying old volumes all the time. Well, let's set up a policy. And this is something that brings you to the initial phase. When you do the discoveries. This is where some reaction behavior that you had in the past where you were fixing things that they are broken or inefficient, how you create new ideas for ensure that these problems didn't happen again.
And I think these ideas go mostly to the platform team.
Hey, let's ensure that we have the proper platform to ensure that we are later not cleaning all the time and we reduce the toil and we focus in things that bring us more value.
I think this is my point in what I'm hearing. What is the superpower? Well, the tooling allow you to do a lot more than you do only with humans. For example, something that we do is like when we want to launch a campaign, for example, like a graviton, how you collect information from all the teams. We have been talking with companies that have like 400 teams running on their it. Then how you get the feedback from all of them, you do it by sending emails or you have a tool that helps you to collect all this information, provide this information to the AI and AI will be able to give you like the proper direction. What we do now with these, should we tackle this? Are we going to have success? Because you know that only 5% of the teams are able to run in year end, for example.
Why to follow this opportunity maybe don't make sense then this superpower for me is like being able to reach more and to do a lot more and focus more on thinking that is what we do the best.
[00:18:16] Speaker A: Yeah, I think, I think you raised a very interesting point because I am very familiar with this.
Especially like AI is very good at processing like a lot of raw data and then being able to provide an answer if especially you are able to provide the context. Right. So what you exactly said is like you have, you can provide the context of your company, the situation, the systems, whether or not you can do certain changes like ARM and changes to Graviton and then you have the data. Like if you optimize to Graviton you will save 20% but then you have the other input that you are providing. Like from a company perspective which is like all our like only 5% of our workloads can run on an arm. So therefore the savings could be like very minimal. But for example if you have let's say the previous GP2 to GP3 update and you can directly run this and it's way easier and a lot of the company will benefit, then you can prioritize this update to the graviton one. So it's all about having the Context and if you can have the help of tools and the help of AI to make you make better data driven decision, as we have always said in Fino, right? So you base the data, then just track the context using or helping with AI and then you take the decision with all these information like, okay, I simplified a lot the concept, right.
I think that's a superpower, definitely you can have. Or the problem that I have been seeing is you have a lot of Data and in FinOps there is a lot of data billing, data optimizations, all of that. But it's like, okay, how do I apply this? Where do I start? What is the most important thing that I need to do? And probably given the situation that we have right now, you can get help of the AI or external tooling and then, okay, this is my context, this is what I need to do, then this is the recommendation. This is how much I will save, this is how much time I will need. Because you also have the human effort that we mentioned. That's also a cost that normally is like, yeah, this is magical, then this implementation will be automatically done. Right. And this is something, a cost that we need to do. And I think that's a great superpower. And Dustin, maybe I want to highlight a bit on this because Dustin, from your experience, what's the superpower that you have seen with tooling and with these new approaches from FinOps that you have seen maybe in you and your clients?
[00:20:56] Speaker C: Yeah, so I'll talk about the superpower before I talk about that. I will say I think the risk to taking advantage of that superpower is the inability to define what good looks like.
So there's a. I don't think people that are not involved in FinOps appreciate all the level of detail and questions that need to be answered on what good, what our standard is for a certain environment or a certain service and the failure to define what good looks like really creates uncertainty, which represents financial risk.
But I will say if we come back to, there are some Azure PaaS services where one of our lead, our lead optimization analysts and researcher has been experimenting with uploading usage data right to Copilot.
And Copilot has done analysis faster than he could have, but it's also done deeper analysis and it suggested some provisioning changes that we were able to implement and see massive savings. And so that's just, that's one example. We're continuing to experiment, but I think that's a, that's a big opportunity area.
[00:22:04] Speaker A: That's super interesting. I think definitely what you said is putting things in perspective and being able to address that and to capitalize its superpower.
It's crucial and I want to be helping the people to take a more practical approach. So maybe I will address this to both of you.
If you know we've been discussing about the proactive to the reactive to a proactive approach. So if you were in the position of these finops, maybe I will start with Ernesto and then with Dustin is like if you were like a finops in a company and you are seeing like a lot of reactive work and then you want to switch to a proactive, what is the very first thing that you would do to change things to make things more proactive?
[00:22:53] Speaker B: The first very thing is like ensure that they have a committee where we make decisions.
Because at the end all the reactive is something that you measure because it's already in your build probably and this is something that is going to be there. But the most important is ensure that we do things before these things are happening. And this only can happen by conversations. This only can happen if we see it and we look. Hey, why we have. For example, I remember a project that I worked like lot of years ago in United nations when I was leading finops there and we had like a big infrastructure full of tomcats and we at some point we were like thinking like why we have all these tomcats here running where we can really.
We don't need all of them running 24 7. We can damage scale to zero. And then we realized that it's really easy to dockerize that and move it to cloud run. This case was running in Google cloud. No. And the impact was massive. Was massive.
It's like 80 times less cost because most of the tokens were never used. They have like very stationary use then for going into into proactive approach. It's very important to have discussion with your team and having engineering in this discussion but also having other people as well. No, like, like lighting the road and the path where we want to go.
[00:24:17] Speaker A: Definitely. Definitely.
[00:24:19] Speaker C: Yeah.
[00:24:19] Speaker A: I think everyone should be involved especially like, you know, mostly most of the times leadership, finance also because finance will probably have insights that you as an engineer won't have. And especially on the commitment. I've learned probably more from commitments and rate optimization, from finance and procurement and probably from other engineering people because they have a totally different perspective. So definitely learn from others and maybe Dustin, so from your side, what is the very first thing that you would do if you were in that position if you were a finops and you want to Change things about me more and more proactive. What's the very first step that you would do?
[00:24:57] Speaker C: Yeah, I think you need a quick statement that motivates people to be interested and also assuages concerns they might have about risk.
So the reason you need to motivate people is because you're, you're promoting, you're, you're proposing a structural change which takes work.
It's, it's always, it's easier just to fix things after the fact, but you have to go back and put in the work. To change the business as usual takes effort. But a proposal like this, okay, I'm seeing patterns of waste, that's not the best way to work. And if we can change that, we can reduce wasted spend. But we can also reduce the amount of time I take from engineers to distract them from their development work.
And we can start in a low risk environment like dev, so we can implement this change with minimal risk.
And if we're successful, we can move up into higher environments. I think that's, that you can say that in 15 to 30 seconds. And that I think can help people be more willing to participate in the work to change things.
[00:26:05] Speaker A: Yeah, and I think like, like from engineers, like we are always looking forward to have to do a test especially if we have like the, if we are provided the information and the tooling and we are already like if you explain something to an engineer and it's like a technical challenge and you can test it on dev especially or in your local or whatsoever, like yeah, let's try that and let's see the impact. Right. I remember like for some of my first kubernetes optimization I was in dev, I was like, yeah, let's test it here. There is no issue tearing this down or moving things. So definitely a start over there. Also as mentioned, treat it like a technical challenge, especially in FinOps. Like we share a lot of theory and a lot of the practice, but most of the finops applied in automation, like it's a technical challenge. It's just another tool, another manifest that you need to apply another infrastructure as code addition that you need to add. So it's a technical thing after all. It's like, you know, you set up your network, that's fine, but you also can have like an optimizer of your network, let's say. Or if you, or your data or your billing, it's another component of the technical side. Definitely.
[00:27:19] Speaker B: I would like to go a little bit deeper on that because I really like this motivational part because this is the one that really touched the human component. All the rest is technical but. And also we know that everybody have different motivations. One engineer will have a motivation to go with cool tech. I want to have something like super resilient. I want that this go fast.
If you talk later with finance probably is going to have another motivation and if we talk with someone with product will have probably a different one then how to motivate them. I think it's important to provide visibility to everybody to allow that the people that is participating in the process feel a little bit as well. He said I'm being able to hang this medal like hey, I'm contributing to this process. But also it's also very interesting point because we are discussing internally a lot in this topic and my co founded Julia recently proposed. It's like hey, why if we align this motivation with values that is something that is common for everyone in the company. Why we align social corporate responsibility with the savings that we are achieving inside finops. This is something that it was to explore and I think it's very interesting because on this case you can be as an engineer collaborating with a coach that you are familiar with or you are interested in, then this will move you forward to do that. But company also will be very interested in hey, if we can feed also the social corporate responsibility budget with some of the things that we are saving in the other side, this is also interesting for business.
Then there's a lot of ways of motivation to explore. And I think it's like a field that was not explored so much.
We did very few steps with a leaderboard and trying to create this visibility. But we want to go one step further than that. And I think it's like really interesting topic maybe for having another full conversation about that.
[00:29:17] Speaker A: Definitely, definitely, definitely. I think you brought a great point. I mean like the connection probably so highlighting on this. I think when we talk about finops and the blockers and all of that and it's always about executive buying and executive buying to get to a TV this normal you come from a technical perspective and you may not be, you know, talking the same language. Maybe executives, they are not that interested into the technical details and they are more into the business transformation, the business value and all of that. And what you highlighted is like what is the values of the company is one of the core elements of business. And on the responsibility side as well. It's like if you are impacting things that are shown in let's say in an investor call in the end of the quarter, you are probably like going to have executive buying if you are impacting those numbers that they are showing in a call with a lot of important people, definitely you will have some buying from the executive if they are looking better. So I think you brought a great point especially you know, driving the, the green upside and the, the responsibility side. Okay. We are improving, we are reducing our, our carbon footprint. We're improving the, the, the resource usage and the company is better, let's say for the environment and for the practice. So I think we definitely need to dive deeper into it. I think you as a finance practitioner, you need to know what's the driver, what's the thing that is motivating people. Find that and adapt it to each of your conversations. If you are talking with procurement, maybe they are having better contacts with the vendors and achieving savings for the company. But for an executive, it could be establishing better values for the company and attaching those to the whole practice. So I think you brought a great point. I definitely want to dive deep more in future episodes.
So definitely want to take a look and maybe I wanted to maybe translate this to Dustin. So what do you see in leadership or in companies like KPG or companies that you would be collaborating with? What do you think are these motivations in terms of values that you know, can drive real phenops buying and real finops change on, on the org basis?
[00:31:42] Speaker C: It was interesting when Ernesto was talking about values. I think about the internal values such as collaboration and integrity and teamwork and I think about recognizing, you know, your stakeholders for, for good work.
So that's what I'm, when I talked earlier about running experiments, we're working on increasing the level of recognition that we give to stakeholders who work with us.
But I think about if you have a good success story and it involves collaboration across multiple organizations, that is something that I think will connect with leadership on how the team is working together.
Of course, some of the things we alluded to earlier with perhaps better energy efficiency or better financial performance, that, that goes without saying that that is a key part of leadership messaging.
[00:32:39] Speaker A: Definitely, definitely. And you know, being able to translate into the concepts and the ideas that are more, let's say technical or more abstract into what drives leadership is one of the core things. And maybe I can, you know, bounce back this, this question to Ernesto Witz already, but this, so from your perspective, like you, you are a CEO, so you come from, from, from a leadership. So for you, what is the main goal or the main value that you want to achieve with, with the technology and what you are doing with. With. With the company and you know how this value that you drive can do better Phenoms. You know, like basically the question that you said, you know, how you can apply phenoms to drive better values in case. In this case in. In Dr.
With the use of technology.
[00:33:33] Speaker B: Well, this is a very long question in the sense, like there is a lot of things that you can help to bring the better values. I think here what we are trying to do is normalize this finish thing and make it like another standard that we have in our cloud operations. The same that we did with DevOps, the same that we did with checkups. And now we are going with finops now. And, and in terms of values, I mean, of course it's important that we are efficient and we ensure that we spend the money efficiently because then we can maybe invest more in people that. In machines that they think is very important most with the times that they are coming now.
But also it's interesting like opening this door to this thing that they was mentioning because we're not talking only about green ops. We are talking as well, like what do you want to do with the savings? Because I've been in some conversations with some leaders that they say I'm inside the budget. Then why I want to save money here my budget is this one. I'm fine. No, but it's like, well, it's the same that I say to my doctor. No, when I visit him, it's like, hey, I'm 100 kilos. I'm fine with that. This is my budget of weight. But maybe I can do far better. No, then if you motivate them that there is another thing that you can do and you can redirect some of these costs, some of this money flow to somewhere else that helps the company in other fields. For example, I don't know, you want to help refugees or you want to help with cancer research or some other things and you facilitate this to the organization. I think this is also another motivation to go behind the values. I don't know if I answered your question, but this is the concept that they have in my mind, like running. And I think this is some of the things that we should consider and also from the engineering point of view as well, one motivation is really have the deep understanding that this is another metric that we are working with. It's not just throughput, it's not just compute.
We also want to be efficient. It's the same when you build a car and this car consumes like so Much and probably people are not going to buy it. I remember some old cars when I was very young were super cool but that was not the most efficient one. Then this sense of efficiency that you as an engineer you're building a piece of software that is also efficient. I mean I guess this increase as well your recognition you are not just doing cool tool but also is very efficient. Then what else? What's next? No, it's a new challenge let's say.
[00:36:19] Speaker A: Definitely it's like the idea of the you know, building the Porsche or building the Toyota. Like the Toyota needs to be efficient, needs to be on budget so that it can be on volume. And I think especially like if you talk to let's say low margin, like business kind. So let's say airlines, supermarkets, all of these that play on very like thin margins that every cost is, you know, you see like for example, I recall like Ryanair's case using carpenter and all that and spot instances like was the saving of the savings of the savings. It was like yeah, super into the lines of the companies. Like you see like the spirit of the value of the company in that approach. Like you see, you know, trying to bring the lowest cost to any level and then you know, you see the engineering being able to apply that into that case. So you know, it's also like it can be always like bottom up that you see you try to translate that message to the leadership but also the leadership, once they understand they can easily translate the message from the value of the company to the bottom so that they make application more efficient if they are. So like for example the supermarket airlines is like we want to make it so efficient. So then let's Translate this to FinOps and let's translate it so that engineering and able to efficiently process that and make it quick and make it, you know, consume fewer resources so that it's, it's more, less expensive for, for the company. Right. So it's like it's a super interesting or like all the motivational thing the, the leadership, the company side is like because there are like super, you know, gray lines about. There is no like automation that you can do with persons.
It's definitely a super interesting stuff.
[00:38:09] Speaker B: Yes it is.
[00:38:11] Speaker A: Maybe continuing on of this more philosophical but I love the idea of the vision is like Dustin, from your view what is the evolution of FinOps will be in the next coming days? Especially with all the changes that we have seen in the couple years. What do you see finops evolving these days?
[00:38:37] Speaker C: Well, yeah, I certainly hope it evolves more into an enabling function as opposed to the central function, doing the work.
And I'm talking specifically about enabling the organization to shift cost efficiency left and equipping teams to do finops work in the context of their daily operations. I think that's a big and identifying, identifying tooling that can to let them achieve that. So that's, that's a big step forward, I hope to see.
I also, I do think that this question is part of what we ought to be thinking about in doing FinOps work. One of the things I like about FinOps and this here my answer gets a little more philosophical, but I like the fact that continual improvement, continuous improvement is part of what we do.
And I think as humans we, we are made to always want to get better, to develop.
And so I think part of our job in FinOps is to continually challenge the organization.
Here's how we can get better. And so if you come back to Ernesto's example, yes, you're under budget, but if you will get serious about this, you can stay under budget and create more innovation capital. And there's lots of examples like that. But this idea of finops as a stimulus to challenge the organization to get better, whether that's doing new things or doing the same things in a more excellent way, I think is critical. And you know, some people call this a growth mindset, but it's the idea that we should not stay stagnant, we should be continually looking to get better. It's part of who we are as humans. And I think Finops is a great field to work in, if you like that philosophy.
[00:40:28] Speaker A: I think that's a very good reflection.
I think especially it aligns with being able to constantly improve, to always have something to do, to not have a finish line. You are always in the search of the next optimization, the next improvement of the process. So it's always ongoing. And I think business are a life thing. So they are always improving, they are always changing, there is always new challenges, technologies evolving, etc. So I am super, super aligned on that one and I want to get Ernesto's one because as a founder of a company, as a leader of a company, how do you see the evolution of phenoms, especially from your perspective?
[00:41:12] Speaker B: My perspective, I think that we have now a great, great opportunity because the use of AI for PhenOps is super important. And I'm not talking only in LLMs that can be very good to collect all these structure information and put it in a structured way to translate what we want to the machines, but also how we do inference how we can collect like a lot more data points to being able to generate like more idea for finops. This is really important and this is something that I've been discussing before with colleagues and I think the future is like ensure that when the companies are talking about finops we don't face what we are facing today. That you go and talk with the companies like no, we don't have finops or finops is a role that the DevOps people is doing. No one do that with security. No one can think nowadays to have anything in production without the security team looking after it. I think in finops we need to go in the same way and ensure that it's really a part, an integral part, a part of all the cloud operations. No, I repeated before and I would like to go this way again. Like the same like cloud ops licence. DevOps people, Seco people, FinOps need to be there as well, need to be part of our pipeline and we need to ensure that when we deliver something we are not just wasting money.
Definitely.
[00:42:36] Speaker A: I think that's the whole idea to integrate it into the part. I think making it, you know, part of the process, make it as automated as possible, as simple as possible because, you know, everyone has enough on their plate is, you know, changing the will change how business will look to FinOps and making it, you know, embedded into the corporate values and the corporate process so that we can achieve better resources, better results and you know, give money to better initiatives. Hiring more people, making the business more efficient or doing in more initiatives that can help others.
Well, I think that's like a very good point to end up the reflection. I think it was a pleasure to have you guys today. I think we had touched a lot of very important people and we share a lot of very important reflections. So Ernesto, Dustin, it was a pleasure to have you here today.
[00:43:32] Speaker C: Yeah, likewise. Thanks for the invitation, I really enjoyed it.
[00:43:35] Speaker A: Yeah. So thanks. And we'll probably dive deeper into more of the. I want to really dive into the more leadership motivational part on the next episode because I think it's a wonderful conversation to have. Maybe we can bring some other leaders to exchange some opinions. And that's all been all for today. So thanks again for being here and see you in the next episode. Bye. Bye.
[00:43:56] Speaker C: Bye. Thanks.
