Episode Transcript
[00:00:00] Speaker A: We are going to talk about what AWS updates from Rayman matter to you. We are going to bring you Luik Fournier from AWS to explain us what does matter from from Rayman and from these last AWS updates and how to plan your first days of 2026. If you would like to join us, please welcome to this episode by Finux Weekly. So let's get to the news. If you have to like ignore everything else and just pick the top three announcements from Raymond for the FinOps team to follow.
[00:00:35] Speaker B: What are they of course the first one will be database saving plan because I would say everybody were looking for that from five to six years and this is really working backwards because I would say this was expected from long time to make the more agile compared to the reservation especially on database to cover the serverless and also some database services were not I would say covered by reservations.
The beauty of this release also is in fact for me this is great for customers. It's more maybe a pain but in fact the fact that we are trying to force the customer to modernize the their infrastructure to be able to use it.
I would say the sherry on the cake also is for the first time we have now a tool that can guide the customer about how they can save and how much they can save immediately with saving plan purchase analyzer integration from the release. And this is the first time that we have I would say for new saving plan reservation features that kind of tool already released in the same times. Great. I would say collaboration between database team and Insight team. So that's also great.
Yeah. The second one is Compute Optimizer Automation. Why? Because I think this is I have not the roadmap in mind, but this is the first one and it would not be the latest.
We are in fact one of the pinpoint that we have with the customer when we are optimizing their workload and we are helping them to discover what are the waste, the main waste. And very often you have 100 or 200 EBS volumes that are unattached because sometimes the devs are DevOps are forgetting to delete the EBS volume when they are deleting the EC2 instance and the customer always like oh, how much time I will need to understand if I can delete.
Yeah. So now the ability to automate and with a great interface the snapshot and the deletion of this unattached volume is absolutely great.
You have the ability to roll back, you have a dashboard to monitor the progress and it's not only I would Say snapshot and delete the unattached. It's also the conversion the famous1 between gp2, gp3 or io1 io2 that has been I would say pushed for years regarding this 20% between GP2 and GP3 giving you more throughput and IOPS on GP3. So yeah a great second one and the third one was not easy to. I would say it's not easy to select but my and this is really personal and especially with my use case and my own customer use case.
The sharing prioritizations what we call RI risk I would say prioritization the ability to decide on which account you would like to prioritize the application of reservation and saving plan because for some company that are having a separate business unit cost allocation it will avoid to have I would say some script or calculation afterwards to know to understand how the saving plan arise have been applied because before that it was of course AWS were applying this on the most interesting discount for the customer.
Nevertheless, even this change in prioritization can affect the customer discount because if you change the prioritization you cannot especially get the best discount but it could also ease the cost allocation.
[00:04:54] Speaker A: Yep, that's like. I think that's a very good summary. I think the savings plans sharing it's also something that could be very useful for for large like organization. I've been like reading a lot like previously on like the complexity sometimes to you know be able to share the setup that sometimes you need to to make so that the sharing like works great and now it's like a very like mostly simplified and that's like super super helpful especially for. For large enterprises. Right?
[00:05:25] Speaker B: Yeah but it's very hard to because the team have released so many new features during just the week before and during your event that it's very hard to select three but at least that the three that have the most impact for my customer.
[00:05:41] Speaker A: Yeah definitely. And I have to say that I was following like last week's updates and I thought like maybe the feed got wrong because of the to get the Raymond ones. But no the issue was that there was like a lot of new updates from the AWS side and I was like okay, they did it again and.
[00:06:02] Speaker B: To be honest on Friday is not part of re invent they have released a very good one again I don't know if you you have seen it already but yeah at least you spoke about this this morning about the account tag. I really love that one also because that is for all the. I would say the untag usage type and the ability to directly I would say list the account with a tag. It's also great.
[00:06:32] Speaker A: Yeah, definitely that's a good one. Especially for the people, you know, struggling with the governance and with you know, finding the right resources. I think it's going to be a funny one. Whenever it's implemented in the orgs I'm seeing like I was already reached out by Rick Triana which we had a podcast some time ago about implementing this like yesterday. So it's going to be fun the results that they are going to have from, from the clients. But you know, moving on from, from the big headlines that, that we had. What do you think like from your knowledge and your respect experience is the, the sleeper announcement, the, the announcement that is small and that it was a smaller in terms of the, of the engagement but this is very valuable for, for the FinOps community.
[00:07:23] Speaker B: I really love the, and I tried to, to, to, to test it before the.
It was. I think this is something highly underestimated with the customer when I'm implementing FinOps governance.
Of course nowadays the customer are more and more familiar with the tagging I would say with the three pillar process people and tools. But one area where the customer are the most struggling is the forecasting. The ability to project their cost.
Of course they are doing budget in finance, they know how much they will have but having the ability to if they have a seasonality or any peaks during the year.
So these features of the 18 month forecasting with AI I would say in Cost Explorer is I think underestimate in terms of value because it will take your last 13, 8 months of historical data, will check it with AI and will give you 18 months of projections that will allow in fact the team to really forecast and be proactive regarding a raise up of the cost instead of being reactive. So yeah I love this to help the customer to be better in terms of forecasting.
[00:08:52] Speaker A: Yeah. Now.
I think also that forecasting is one of the. Along with budgeting which they are you know sometimes very related are the two topics that are the most difficult in FinOps. That's my personal opinion maybe with unit economics. So you know having more tools and having more ways to make this easier for, for the, for the finOps practitioner is, is always helpful. Right.
So we wanted to you know, now that we have the, the updates and now that we have seen what could have been missed and how you, you provided this, this good tip. You know, let's, let's imagine that we are a finops practitioner and it's the Monday morning or maybe now after re invent and this new batch that we had today. So what do you what is your step by step checklist for for a FinOps team that should execute let's say in the first 30 days of 2026.
[00:09:52] Speaker B: Okay, so my advice will be trying to I would say save from the first day. So of course regarding the database saving plan especially for everything which was not covered until this release. I mean DMS serverless database documentdb for instance which give you up to 35% opportunities I would say we have no tools yet to I would say to help I would say it should come and do not hesitate always if you have this kind of feedback especially for Insights team because Rick and Jimmy and all the team love to receive feedback, they love to work backward and all the features that they are implementing are coming from the customer feedback from the field.
But yeah, having the ability now from the day one of this database saving plan to check with your purchase analyzer and check how much you can already save. It's great. I love to say it's important before committing to always right size in documentdb. I think the right sizing is not the same priority as it could be for LDS database or whatever. We have no tool yet I would say to help directly as Compute optimizer to help the right sizing then as soon that is being done also check if what are the statement of your regarding your rds what is the statement between what is already reserved what is still on demand and check if you can upgrade to seventh generation or onwards and especially an advice if you are not for all your managed services on Graviton switch to Graviton. This is not linked to a new feature. Even we have Graviton 4 now but it's must to have it's a no brainer or your managed service should be on Graviton to benefit from better price performance and better sustainability.
Then of course have a look if this has not been done yet have a look on your EBS volumes if you have a massive EBS volume that you have never took the time to delete check the automation from Compute Optimizer and because we release.
It has a cost so now we are detecting the unused NAT gateway through Compute Optimizer and also we release the regional unite gateway. So before you have to get one NAT gateway per az with a public subnet and now you can have a regional one.
So it's great if you can transition from your 3 nat gateway if you are 3 az to 1 only regional nat gateway that's also good advice. And then I would say let's go to the week two where we could now we have reduced the waste. We are right size. We have made commitments we can enhance the visibility and the control.
So if you have 20 pairs I would say up to 20 pair more than one at least. And you were struggling because you have to switch from one to the other to check. You can now aggregate up to 20 pairs in a unified view all your pair account build. So that could be also interesting to save time custom detection we have extended the dimension for the proactive alerts. You have now 24 hours a real calendar and also an automatic I would say manage monitor so you don't have to create your custom one. We have now one super intelligent that could also trigger that you could have missed with your custom ones. So yeah update to the new manager customary detection monitoring.
And for also you have I would say depending if you are using already cloud intelligence dashboard. If you are not, you have now the ability and it was just before reinvent the ability to create some dashboards.
On cost Explorer. You can create dashboard to highlight some specific KPIs and you don't have to each time reset up your granularity. You can save it into a dashboard and have a full control.
And then for the week three of course I talk about that check on your 18 months forecasting to prepare 2026 based on the last 38 months.
Also check if this is not done your sharing preference on the arise and saving plan because maybe you can decide to prioritize some business unit or and also check your cost efficiency score.
It's a longer I would say.
I think Victor can share the blog, but it's a new way to estimate.
I think every company has it own way to estimate the the progress and how much you are saving and how much you are going more efficient.
But we have tried at least not me but the team has tried to create kind of standardization to evaluate the efficiency score and to allow you if this was not done to directly having comp cost optimization hub this value to estimate your progress.
And last but not least and this is very important to add some guardrail regarding AI.
AI is great AI for finops. Everybody is talking about AI for finops. That could help a lot. That could speed expedite many things.
But we are all aware that AI is not yet perfect compared to human. And I mean the accuracy is not 100% yet.
So it's very important that when you are activating usage of AI for FinOps that you are always capable to investigate what the AI can do or what the AI is doing to be sure that what you will do with AI is fully accurate. So, so pilot test and so what are the new we have enabled so we have Kiro Q developer which is becoming Kiro.
So it was a bit before Re invent but for instance we have integrated this with many API and the pricing one for instance. So instead of using the price calculator you can if you have an architecture to be done use it. I think I wrote a blog about that. It was absolutely nice. The billing MCP server integrate customer development workflow and also of course if this is not yet done if you are using Kubernetes have a look on the granular EKS cost attribution for your cost allocation.
[00:17:21] Speaker A: That's great also yeah, that's like a lot of job to do for our fellow Finns practitioners.
[00:17:30] Speaker B: But you know it's always what we are saying to the customer. It's not a one time effort, it's always a continuous effort. And what is important with all this information because you told us that it's even hard for you to follow and to be up to date. What is super important is for the FINOP practitioner to educate their team because if they are trying to do any everything by. By themselves it's very hard. So they need to empower the.
I would say the. The culture of, of custom awareness inside their organization. And to do that I think I encourage to do like a finops reinventory cap where they're explaining they are doing demo about this new feature and also not yet but soon we we will have a I would say workshop available.
We'll have also game day. So it's very important to find some safe environment where the customer can practice and enable the dev and DevOps to understand this new feature to be able to use it.
[00:18:40] Speaker A: Yeah, yeah. I think the main issue is that they will probably have to do a recap every month or something with the amount of updates that AWS provides as you mentioned like for me like I dedicate some of my time to follow that and you know I do it for trying to help the community and it was difficult for me to try like everything if I you know, miss something or whatever with the amount of updates that they are. But you know I definitely recommend people to investigate. This is probably like a year long list of work if someone is like alone in their team and all of that. So you can do step by step. We have highlighted all the priorities so I think this is a Very useful resource for plan.
[00:19:28] Speaker B: And also there is the Focus. I forgot to mention it but the Focus 1.2 has been released. Even Fina foundation has released the 1.3.
I have no insight about when we will release the 1.3 in AWS, but 1.2 is available and it's great if you are multi cloud also to switch to focus.
[00:19:50] Speaker A: Definitely. No, I think it always takes time to introduce the new updates especially with all the changes that we have seen and especially with all the now that we are focusing more in the cloud plus and all these it requires a lot of new columns and all of that. So I would like to see the next implementation of the focusing initiative both on the new new version and the implementation by the AWS side.
But you know we have talked a lot about the. You know, the new releases, the new updates, the new stuff. But you know we think and we see that new services like often come with new pricing models.
So from your experience do you think there are any hidden cost implications or trips in these new launches and launches that people need to watch out for?
[00:20:43] Speaker B: Yeah. So I would say for database saving plan this is the feedback that I already received from my customer.
So for instance some customer have very small database like T series. That is because we have only 6 generation on T series so this is not yet covered by database saving plan. So up to upgrading to a 7 generation meaning a C7 or could be more expensive that what they have today.
Also some customer would like to do hybrid between reservation and saving plan and we have not I would say official tool to do to do the math because of course this is the trade off and we are trying to.
The goal is to push the customer to adopt the saving plan for their database. Why? Because even if they are doing individually they can using reservation they can get more discount. That's a reality.
But if you think about the cloud is the agility because every project can drift, every project can switch or increase or decrease and having the ability to change the family, the switching from provision to serverless or serverless to provision is I think that we should not estimate only the direct cost and the direct discounts, but also the agility.
And when we are talking about agility is really the conversion between the owners all the services that we have into the database someone willing to change from MySQL to Postgres or from serverless to provision.
That's really important for Elasticache also the Valky conversions.
I think that's something super important to remind we have elasticache that will start start to extend its support. I think it's a end of Feb or beginning of Feb, I don't remember exactly the date.
So there is two things here. First, if you migrate to Valky, you can get to 1 to between 20 and 30% of discounts compared to Redis.
And also you are getting a better latency, better performance and this is the only one that is covered by the saving plan. So if you want to get your elastic cash covered by the database saving plan, you need to build Valky.
[00:23:25] Speaker A: Okay, that's. That's a good tip.
[00:23:27] Speaker B: Yeah. Yeah. I mean although other than this I think the. The.
The combination between ARIS and.
And saving plan database saving plan I think will be be the challenge of many customer.
Because also the other challenge is when you have a database and if you are checking the price between fifth generation, sixth generation, seven generations you can have for instance you can see that the seventh generation can be more expensive.
And what we sometimes promote for compute is not the same story for database. So we can always say to the customer yes, it's more expensive but you will get 30%, 40% gain into efficiency. And you can maybe downsize your instance type.
The drawback with the database is the memory because memory is an important part of the database. How it's working, how you are loading data in cache. So downsizing the database memory because this is linked with the number of VCPU is not always possible. And so this is why some company will say okay, moving from the now from my reservation on six or five generation and moving to seven generation for getting less discount with the database saving plan and also paying more expensive. My the.
The instance type could be first like a broker. But again you need to add the agility inside the equations. And I think this is what. What matter the agility that you can. You will have to. To switch.
And again also very something that is important when you are committing for saving plan.
You have seven days to roll back. So after 24 hours you can already seen in cure the benefits or not of your saving plan. So if you are seeing that finally this is not what you aim for. You can ask to. To roll back and return to your previous reservation.
[00:25:48] Speaker A: Yeah, that's. That's a good one. I think like you have. You know you have a kind of weak free trial even though it's not like super free. But you know you have a week of trial to.
To see if it's working like savings plans. And you know reserve instances is a huge commitment like from the. From the cash.
Let's say the Finance perspective. So it's good that you have some margins, right. That you can play available like okay, let's test it out, let's see. Cool. And then let's decide, right that you can, you know, make. Make some simulation. Because otherwise, you know, on paper sometimes it looks good, but then when you apply it, it's not working, right?
[00:26:29] Speaker B: Yeah. And it could be also the good occasion for those who were, I would say not going on serverless because of the lack of saving plan or commitment or discounts.
It could be the right time to reevaluate the benefits of serverless.
Especially because, you know, on Aura you have the ability to get hybrid.
It can be a hybrid so you can have a provision writing and serverless reader. So I think sometimes when you have a LDS multi az you you are paying one active, one passive with or you are also reducing the operation need. So you can reduce the resource impact and the operational effort for your resources going to. You get already the storage which is already multi az.
And you can use serverless, but serverless, you know, as already it's also dependent on if you have a 24 hours running database. Serverless could not be the best or scaling and scale out is not an option because. Because you have full run all the time, 24 hours, seven days. But considering R with the hybrid setup could be also something valuable now.
[00:27:55] Speaker A: Yeah, definitely.
I think it's very, very useful. I followed like when I started in, you know, in AWS development.
I think it was 2021 or something like when I saw like the first upload as serverless and I was like, yeah, this is a great idea. Because probably databases are one of the most like, you know, continuous workloads that you can have in cloud. And you know, having a hybrid model able to fix that, it could be. It would be great. So. But they want to. You know, I am a very handsome person. I am many trying to get real numbers to people.
So maybe we can go through a real scenario.
Maybe if a client we can go through like imagine you are a client and it adopts announcement. One of the announcements that we have mentioned. How does the math change compared to the previous status of their deployment?
[00:28:49] Speaker B: Yeah, so for instance, taking an example of a large organization that is using a complex database portfolio. So they are using MySQL for their core application, DMS for the database migration, Neptune for knowledge graph and documentdb for content management.
Monthly spend 9500 under the old model. So before the database saving plan they could only get reserve instance coverage for Aura, MySQL it's a provision one.
So we can say and this is average figures to make it simple, they save 35% on 5000 and bringing that to 3250. But the remaining 4500 across DMS, Neptune and DocumentDB were still on demand. Okay, so the total monthly cost is 7750.
Overall the discounts globally compared to the 9500 is 18%.
Now with the database saving plan you are you can cover the all the services.
So you get up to 35% coverage. And we to make this simple we will drop to 6,175 monthly. So now the globally the saving across all your database service will be 3325.
Meaning that it's versus the 1750 which were give by the aris. We are talking now about 18900 additional saving per year.
So I think this is really a great additional saving that you you can do. Because with the previous version only aris you have a monthly cost of 7750 and now you are going to be at 6175.
So globally.
This is why I put database saving plan as a first thing to check because you can start from day one after the release to have directly a cost reductions on your on the services that were not impacted. Even without talking about upgrading to the new generation or whatever you have already. Services that put potentially can be covered from day one.
And as you know this is not because you will you can take how many saving plan that you want. So of course more you will take saving plan. The return of investment is decreasing depending how you are sharing the priorities because the first one will get the better discount and then we are of course this is decreasing in term of roi. But really have a look on on this. Especially if you are using those documentdb Neptune DMS that were not covered before and serverless.
[00:31:51] Speaker A: Definitely. No, I think that you always have to be up to date with the updates. I think there is now that we have more coverage on new services. Sorry, it's just being up to date and being able to check the current spend and being able to adapt our commitment strategy to. To feed those right. It's. It's a great, great, great update. So.
[00:32:14] Speaker B: And, and yeah, and again I would like to say if you have any feedback, do not hesitate.
So if you have a account team on AWS TAM account manager, don't hesitate to tell us what is your feedback or even you can contact Victor on LinkedIn or any can give it to me or directly Contact me on LinkedIn We. We need to hear from. From the field, what are the good. What are the. What could be improved on this database saving plan? You know, I would say 99 of the feature that AWS release are just coming back from the. The what we call internally the pfr, the product feature request that are pushed by the customer. We are working backwards from the customer. We are customer obsessed and this is super important.
Even if you have a feedback, it will not be released in two months. Of course because there is everything that is already maybe scheduled for the next six months. But if this is relevant and many customers are pushing the same feedback that can also be more faster. Take into consideration generation is improved by the team. So it's very important to not say okay, that is not good and keep it for you. Share it, share it with us.
[00:33:34] Speaker A: Yeah, yeah, no, I'm. I'm very familiar with it with the Amazon like with the AWS feedback cycle and yeah, feel free to reach out to Luig or to me and I can redirect people.
Yeah, of course, if you have like a Deckout team or, or whatever so that we know they can help you on having this feature. Right. So okay, so we've been talking about like you know, 30 minutes approx and we don't know a lot about you Luke. So you know, you are now a tab manager in aws. But you know, take us a little bit back on, you know, what were you doing before? What triggered you to join phenoms and AWS?
[00:34:15] Speaker B: So, so I start my career in IT in 1999, in the last century and I'm nearly 45, I'm based in Geneva and I'm dedicated I would say working on the international organization.
And what I have done before AWS had I would say very several roles in infrastructure. So security, networking system, database. I even work for so many years on Mainframe, S400i series, IBM.
I've been also middleware integration on Java, Tomcat, WebSphere and all these nice things and then also database Administrator on the db2 Oracle.
So I had a very nice career. You know, we have to reinvent ourselves myself in IT many times and learning and be curious is always great.
So I was freelance for many years. So when you are freelance and you might know what it means if you want to stay longer with the customer, you have to be more than just a database consultant or you have to be more.
And then I also transition to business also then I have a role of business analyst and because I realized after many years and more experience that it I would say infrastructure is just a tool for business and if there is no business, there is no it.
So it's very important to understand and to speak the same language of the business.
The ability to speak with cfo, with CEO, with I would say executive, but also with the business unit to understand clearly their needs and provide the solution that really fit the needs and will solution what they need instead of looking always for the last generation of all the best solution in IT perspective and why I would say I joined AWS three years ago.
@ the beginning I have no clue about finops, to be honest. I was passionate by fine finance, I was passionate by sustainability and also I'm very customer obsessed.
I spent many years in support area. I was director of operation and support. So it's very important for me to always see my customer happy. And I find that what is one of the differentiator of AWS is this customer obsession and the way that we are as a term were empowered to help the customer to reduce the bill. It's quite, I would say when you're spending time with your customer at the first time, always the customer reaction is this is quite nice that we have someone from AWS helping us to reduce our AWS bill.
You know, it's not common and so I really love it. And also I start to work with cloud intelligence dashboard team to develop the features. Features with Yakov and Yuri and yeah, more I was working on finops. I discovered that it was a collection of everything helping the customer to be more sustainable, which is super important when you have kids and you want your kids to grow in a better world, I would say.
And also you want your customer to be more efficient, to be more modernized and then of course, course helping them to save. Course. So this is why I would say I decided to learn more and more on this area and trying to become a very good specialist in finops because I think that was a good connection with all my favorite.
[00:38:08] Speaker A: Topics.
Okay, that's a good, like, that's a good story. I think it's, you know, wonderful to have like so, so experienced people. You have a long career already and I think you already have like a long career to continue. So definitely a lot of knowledge in that head.
[00:38:27] Speaker B: Right? Yeah. And I think it's cool because I did most of my career on premises and you know, on premises we have this nice ITSM principle.
And when you were doing a program project, you were, I don't know, you were doing a meeting. Okay, this is how what we will buy. Okay, we will buy this you were buying the hardware, the license and then you were not speaking with the business for three years. So you were just making the day to day and that is working the business opening case time to time why. And you were doing change management.
But the, you know now in this the trend is the chaos engineering. So there is no anymore any change management that okay, we will do an upgrade on your production in three months. So we will do months one we will do dev, then test and prod. Now it's not going like this anymore.
So the agility of the cloud has opened I would say. And you need. When the customer are doing migration we are helping them of course to do the migration.
But in the FinOps governance area I'm trying to help them to change the mindset because the first what they are doing when they are doing a migration they are.
I would say usually they are putting every infrastructure the double of what they need because they don't know exactly how even we have tool to tell them okay this is what the instance size should be. But they always feel fear about to not have enough power so they will scale everything twice that what they need.
And so you are explaining that yeah in the chaos engineering trend and the fact that the change are the brooklyn deployment everything is. It might change fast and you need to be agile and the finance reality will be completely different in the cloud compared to what it was on premises. And as I know the two worlds I can really explain the difference to the customer and try to make them aware that as soon they are migrating to the cloud they need to start this important mindset conversions to some people need to manage the cost on premise to everyone need to manage the cost on the cloud. Everyone need to have ownership and responsibility of the cost on the cloud.
And this is I would say for me having known the two world this is clearly something that I would say I can testify and I can really help to implement this finops.
[00:41:08] Speaker A: Governance.
That's great. I think it's. You know governance is one of the trickiest part and you being able to. Now that we are doing like all these and the cloud plus adaptation into finops you also have like. You know now that if that comes sometimes it's like you have a lot of insights coming from. From that world and you now have to see like the evolution of. Of the cloud directly and how phenoms can you know apply to. To both worlds. And that's. That's super interesting. And you know thinking about the. You know always the. The practitioner like if you would like to share one piece of advice to to the fitness practitioners to survive all the. You know the updates that they are like every week I am posting like I don't know, 10 updates every week. And you know, apart from the reinvent what what it would.
[00:41:57] Speaker B: Be.
Yeah. First what is super important I think of course there are a lot of people that on LinkedIn Victor and also my colleagues Steph and Savannah and who are trying to through the key of cost optimization doing one session per week where they are explaining the new features.
We have also nice blogs that are what we are trying in the blog always we are starting from the real use case. So of course there is documentation explaining step by step or you can implement the features. But having a real use case in mind where we are starting from this on the blog and we are going down deeply on the features.
I think is great. Also if you have not seen it, this is super important. If you can check the session of the team in Re Invent they are available Mostly available on YouTube the session of Rick of Jimmy Liam, the product demo, the roadmap.
Also that is super interesting to see because they are really explaining because they are the product manager. They are really explaining why they are doing this like this. And it's also helping to understand what the PM has in mind when he asked create the prfaq and when he has designed the product that has been released. So having in mind I would say watching this presentation is also great just after Re Invent to understand what is the target of the product management of the cloud financial management team at AWS what they have in mind. They are explaining it very well.
And also again it's super important we talk a lot about the commitment and I'm seeing still some people that are committing but never optimizing, never reducing the waste.
So I talk a lot about database saving plan. But one of my favorite product is Compute Optimizer and you need to abuse of Compute Optimizer because when you in reality what we were doing before manually when I mean when a customer two years ago before for instance LDS was an aura were inside Compute Optimizer when a customer was telling me can you help me to right size my database? You cannot imagine how much time you were we can spend on collecting the logs, identifying the right KPIs, understanding the business how the database and now having something that tell you exactly and not only telling you but also showing you with graph and projection of the future instance KPIs. It's super great.
And of course we already talked about that but yeah, it's a good time now I would say to check the forecasting for 2026.
Check the forecasting, identify where if you have a big race up, identify why and try to I would say investigate and educate your team. That's also something super important. It's good that as a finops practitioner you. You are again I already mentioned it but the education is great because more you will be able to educate your team, celebrate the achievements and you can even implement rewards in your team. Okay, challenge.
So for instance look we have these features that has been released.
The team that will implement it and get the more. I would say the more benefits in the first quarter will win. I don't know, I don't know a nice T shirt or whatever. But yeah, trying to celebrate the achievement of your team and educate them. We have also on skill builder we have a skill builder, we have nice training about finops that Natalia and many Nudit and so on are put together.
So they are free Cloud financial. I don't remember the name exactly, but there is a training on cloud financial management.
So if you can you. Because the Dev and the DevOps of course if you are asking if they prefer to learn about finOps or learning about the new AI features I think the trade off will be done super fast.
But if you are able to explain what is matter because if you are able to see save we will have more budget to I would say spend money on this super nice new techy AI model and you will. I will be able to. To give you more budget to.
Because you have saved on the production you will have more budget to experiment and develop the future.
That will be another no sense of what they are doing.
So yeah, it's very important also to explain to everyone that from the design it's very important to consider finops and why they need to do that. Because if you are telling them just help me to save money. Yeah, it's not my company. Okay, but yes, but if you are saving money you will have more budget to experiment. That could be a nice.
[00:47:44] Speaker A: One.
Yeah, yeah, you need to pull like the benefits on the real side like the incentives. No, it's always with engineering I think it's always about incentives is getting more people in your team being able to develop more features because they cost less so you can have more for. For the same pricing. And I think you, you know gamification as you mentioned and you know incentives in general, everything we talk from.
[00:48:12] Speaker B: The beginning is just tool. So tools are I would say are great but are not Magic. So if you are nobody you are using it it will not apply. So you need to I would say first of course educate the people to be sure that this new feature will be used and then also you can implement processes. What I mean for processes for instance, if you would like to be sure that you will get the most of database saving plan do not allow to deploy any more any elastic cache on Redis and implement the process to be sure that every elasticache will be deployed with Valky to be sure that every managed service will be deployed with graviton and you can do that. I would say with SCP you can control with config you can detect which one are not graviton and then go to discuss with the team. So implement those guardrail and those control are also part of the deal. And and I would say it's good to implement these processes because even you have great tools. If you have to go back to this the same issue again and again it will be a waste of time. So using the tool is great but be sure that you will have the. I would say you will avoid to reiterate the same action to save again on the same topic.
It's very important to implement the Gua Rain and the process is to.
[00:49:41] Speaker A: Avoid that definitely having the you know, policies that are the SCPs the automation in place so that it happens, you know, without you having to take a look on that or having to recheck or having to manually do so. It's. It's wonderful. So thanks thanks for all the the tips today Luig. It was a pleasure to you know, be able to help everyone in the. In the finos space to you know, gather themselves around on the.
On the AWS rename. So thanks for helping us today.
[00:50:10] Speaker B: And don't hesitate if you have any feedback about this.
We are all, I would say all the cloud financial management team in AWS are very connected with the service team.
It will be a pleasure to give the feedback to the team and they are waiting for it because this is always the stress of the PM to understand if what they have delivered is exactly what was expected and how they can make it the more customer happy. So yeah very important feedback Feedback Feedback yeah.
[00:50:41] Speaker A: That'S. That's great. So so try and help the. The team from AWS so they can help you save more and more after after that. So yeah thank you for, for being here today. I hope you had a great time and see you everyone in the in the next.
