AWS re:Invent 2022 - Optimize your AWS cost and usage with Cloud Intelligence Dashboards (BSI304)

AWS re:Invent 2022 - Optimize your AWS cost and usage with Cloud Intelligence Dashboards (BSI304)

AWS re:Invent 2022 - Optimize your AWS cost and usage with Cloud Intelligence Dashboards (BSI304)

Do your engineers know how much they’re spending? Do you have insight into the details of your cost and usage on AWS? Are you taking advantage of all your cost optimization opportunities? Attend this session to learn how organizations are using the Cloud Intelligence Dashboards to start their FinOps journeys and create cost-aware cultures within their organizations. Dive deep into specific use cases and learn how you can use these insights to drive and measure your cost optimization efforts. Discover how unit economics, resource-level visibility, and periodic spend updates can make it possible for FinOps practitioners, developers, and business executives to come together to make smarter decisions.

Learn more about AWS re:Invent at https://go.aws/3ikK4dD.

Subscribe:
More AWS videos http://bit.ly/2O3zS75
More AWS events videos http://bit.ly/316g9t4

ABOUT AWS
Amazon Web Services (AWS) hosts events, both online and in-person, bringing the cloud computing community together to connect, collaborate, and learn from AWS experts.

AWS is the world’s most comprehensive and broadly adopted cloud platform, offering over 200 fully featured services from data centers globally. Millions of customers—including the fastest-growing startups, largest enterprises, and leading government agencies—are using AWS to lower costs, become more agile, and innovate faster.

#reInvent2022 #AWSreInvent2022 #AWSEvents


Content

0.15 -> - Hello, good afternoon everyone.
3.18 -> I hope you all are as pumped
5.31 -> as we are about all that we announced this week.
9.18 -> If there is one topic almost everyone cares about,
14.64 -> especially during these macroeconomic conditions, is cost,
19.44 -> and how to manage those well.
22.74 -> Well, a great management thinker Peter Drucker once said,
26.467 -> "You can only effectively manage things
30.24 -> that you can measure."
32.34 -> After working with several large successful customers
36.12 -> running cost efficiently on AWS,
38.61 -> we've learned a corollary to that,
40.44 -> which is you can only effectively drive change
43.86 -> in a large organization if you're able to contextualize
48.3 -> what you're measuring,
49.86 -> and then share those insights with individuals
52.83 -> and teams who are inspired to take action.
56.25 -> In today's session, we'll discuss how you can scale
60.51 -> and create financial awareness within your organization.
63.87 -> how you can drive cost efficiencies
65.91 -> with your engineering teams
67.77 -> by analyzing your cost and usage data
70.17 -> with Amazon QuickSight.
71.73 -> I'm Rohit Pujari from the QuickSight leadership team.
74.19 -> Joining me, I have.
75.6 -> - Hi, I'm Aaron Edell,
77.16 -> I run the Cloud Intelligence Dashboards team
79.56 -> and general go to market around FinOps at AWS.
83.46 -> Really excited to be here with you guys.
87.24 -> - All right.
90.09 -> For those who are getting started,
92.31 -> Amazon QuickSight is a hyperscale
95.43 -> business intelligence service that you can use to
98.61 -> build data rich experiences for your organization.
102.66 -> We launched Amazon QuickSight in 2016
105.54 -> to support some of the most sophisticated
107.76 -> and demanding sets of customers,
109.83 -> such as Dolby, Siemens, NFL, just to name a few.
115.56 -> And we built it the Amazon way,
117.63 -> which means we can get started quickly
120.57 -> and deliver insights within your organization
122.94 -> and also to your end customers
125.37 -> without needing to manage servers or infrastructure.
129.3 -> It delivers consistent performance at all scale.
132.9 -> For example, Fulfillment by Amazon,
135.99 -> which is an amazon.com service
139.054 -> that our third party marketplace sellers use
142.68 -> to outsource their shipping to amazon.com.
145.86 -> Hundreds of thousands of those users
148.47 -> use embedded QuickSight dashboards
150.42 -> to understand their sales performance
153.33 -> and make inventory decisions.
156 -> And last but not the least,
157.8 -> Amazon QuickSight comes with consumption-based
161.13 -> paper session pricing.
163.14 -> Which means, and which our customers tell us,
166.02 -> it's orders of magnitude cheaper and cost effective
169.56 -> than traditional BI solutions.
171.87 -> So now you have some understanding of what QuickSight is.
177.12 -> To kick things off, we'll first describe
182.07 -> the value of Cloud Financial Management and why it exists.
187.59 -> And then while doing so, we'll run through a scenario
190.62 -> that we've seen commonly play out
192.42 -> in many organizations we've dealt with.
195.39 -> And while doing so, we'll introduce a few personas,
198.24 -> we'll talk about their needs, their desires,
200.67 -> and their wants.
202.44 -> And then we'll introduce and explore
205.32 -> Cloud Intelligence Dashboards
206.82 -> and describe how they help those use cases.
210.66 -> And we also have a demo lined up for you today,
213.72 -> where we will showcase key FinOps use cases
216.78 -> that these dashboards enable.
218.58 -> So you can take a mental leap and imagine
221.04 -> what these dashboards can do for your organization.
225 -> And then we'll touch upon how you can help create
227.67 -> financial awareness using capabilities of
230.43 -> Cloud Intelligence Dashboards in QuickSight.
233.19 -> And finally, we'll bring in our star of the show,
235.5 -> my Mike Graff from Dolby to finish off strong
238.44 -> by telling their story with Cloud Intelligence Dashboards.
241.83 -> All right, let's get started.
246.09 -> Gardner estimates that global cloud spend is on track to hit
251.887 -> $495 billion by end of 2022.
257.79 -> Although that's such a massive figure,
261.18 -> it doesn't surprise us one bit,
262.83 -> given how strong of an enabler cloud has been
265.2 -> for customers of all stripes, from startups, to enterprises,
269.73 -> to small and medium businesses.
272.46 -> But what does give us a pause though
274.62 -> and makes us think we could do better,
277.44 -> is the Flexera study that states
280.56 -> enterprises waste around 32% of their cloud spend.
286.38 -> Well, it's important to note that spending can be good
290.64 -> if it's helping you drive proportionate business outcomes.
294.63 -> If it's helping you acquire new customers,
297.45 -> if it's helping you grow revenue,
299.79 -> if it's helping you stand apart from your competition.
302.73 -> These are examples of good spend.
306.81 -> So not all spend is necessarily bad.
312.21 -> But wasting, most certainly is.
315.69 -> You wanna avoid paying for resources that are sitting idle.
319.77 -> You want to avoid paying for underutilized capacity.
323.94 -> You want to spot architectural inefficiencies
326.67 -> in your applications the moment they creep in.
330.36 -> But then how do you go about differentiating
333.33 -> between a good spend and a wasteful spend?
336.33 -> And more importantly,
338.22 -> how do you then make decisions about where to invest
341.1 -> and where to pair back?
344.22 -> Well, a chief enemy of good decision
347.46 -> is lack of sufficient perspectives.
350.58 -> And Cloud Financial Management, if done right,
353.61 -> is an antidote to that.
356.58 -> Cloud Financial Management is about bringing teams together,
360.33 -> from finance, procurement, engineering, operations,
365.19 -> together to enable principled, data driven, decision making.
370.2 -> It's also about making decisions
371.61 -> that allow us to move faster while being able to,
375.03 -> while being financially responsible for our organization.
378.27 -> And it's about seeing your cloud spend
379.8 -> in terms of the business value it delivers to you
381.87 -> and the customers you support.
385.14 -> So the same study Flexera found that
390.54 -> companies who have earnestly practiced
394.2 -> Cloud Financial Management
395.43 -> are meaningfully benefiting from that.
397.95 -> 67% of the enterprises say that CFM has helped them
402.06 -> grow their revenue.
404.252 -> So how do most customers get started?
408.48 -> Meet Amy.
410.31 -> Amy is the FinOps practitioner.
412.14 -> She's breaming with confidence.
414.63 -> She's just joined the company.
417.69 -> And prior to joining, she was an IT program manager.
421.92 -> She had delivered a few migration projects,
424.269 -> so she's aware of AWS, but FinOps is a new area for her,
429.51 -> it's a new functional arena.
432.57 -> And her company had recently migrated to AWS.
437.25 -> Her company chose to do a lift and shift migration,
439.68 -> as that was the fastest path out of the data center.
443.07 -> So now teams are comfortable with the cloud operating model
446.22 -> and they're ready to take on refactoring projects.
450.45 -> And Amy is now at the helm
452.19 -> to come up with a cost optimization strategy.
457.29 -> So Amy goes by what she knows,
459.36 -> she talks to some of her colleagues
461.58 -> and she learns that she can go to Cost Explorer
465.3 -> and see her spend, so she does that,
468.36 -> she logs into her Payers account.
471.03 -> She can see her spend over time,
473.19 -> she can see which services are contributing to that spend.
476.7 -> She can summarize the spend by services, by accounts,
481.98 -> by regions, by instance types.
484.5 -> She feels good about that.
486.99 -> So as she is getting ready to provide spend update
490.8 -> to her CFO, she's still wondering at the back of her head,
496.47 -> how do I get the next level of detail?
498.81 -> Like, how do I make this information actionable
500.73 -> to my engineering teams?
502.86 -> So let's see how that conversation unfolds.
505.5 -> - So this is what happens.
507.81 -> Our CFO says, yeah, that's good, that's great,
510.96 -> I'm really glad you're able to tell us about some
513.75 -> high level spend and some cost information
515.67 -> about our AWS accounts.
517.68 -> But I need more than reporting, I need a strategy.
522.81 -> We need to lower our costs,
524.58 -> we need to get on top of Cloud Financial Management
527.58 -> so that we can make sure that our unit costs
529.77 -> and our margin are in line with what we expect it to be
533.1 -> as an organization.
535.23 -> So as a CFO, I'm asking Amy,
538.83 -> what will our costs be next month?
541.53 -> What are we projecting it to be into the future?
544.41 -> We need, and we have a new product coming out.
546.69 -> We need to set pricing for that product.
548.76 -> How do we do that? How do we make that decision?
552.9 -> Everyone in my organization, including my engineers
556.2 -> and all of the people who work with me and report to me
560.16 -> need access to this data.
561.69 -> They also need to be able to
563.13 -> be a part of the decision making that happens
566.13 -> here at our organization.
567.84 -> So if I have my bosses who are in the C-suite,
571.86 -> and a hall of engineers who have questions as well about
575.58 -> what kind of product decisions should they make,
578.07 -> how can I address this for everyone
580.86 -> and scale it out to everyone within the organization
583.47 -> and give them access in ways that make sense for their roles
586.89 -> and make sense for the contributions that they can make?
593.04 -> - All right.
594.69 -> So now Amy's trying to parse the request from Martha.
598.14 -> She knows that Martha is looking for
601.56 -> her to come up with cost optimization strategy.
603.93 -> She's looking to price products,
606.15 -> she wants to scale the financial awareness.
608.55 -> So Amy does some research,
610.77 -> and she learns there are a multitude of ways to save on AWS.
615.57 -> You can use Reserved instances, Savings Plans,
619.14 -> Easy to Spot instances,
621.6 -> there are different types and kinds of storage tiers
626.1 -> you could use to improve
628.14 -> and get a better price for your storage.
632.13 -> She also finds that there are potentially opportunities left
635.88 -> to right size because if you know,
638.01 -> Amy's organization had chosen to lift and shift
640.23 -> most of her workloads.
642 -> So she has all these great ideas,
644.82 -> but she's still unclear as to where should she should start.
650.28 -> And rightfully so, 'cause-
652.23 -> - There are so many tools, right?
654.72 -> There's so many places for her to go
656.34 -> and get this information.
658.08 -> And she learned about Trusted Advisor, for example.
662.88 -> She can use that to get
664.26 -> some really good right sizing information.
666.87 -> Oh, and there's all, there's still Cost Explorer,
669.51 -> we don't wanna forget that,
671.52 -> that's a really good place to go in
673.05 -> and you can build reports,
674.13 -> and you can customize the filters and the groupings
677.82 -> and these kinds of things.
679.56 -> Oh, but then there's also AWS Budgets.
681.6 -> So AWS Budgets, awesome tool, I can set budgets,
685.02 -> I can have them auto increase based on percentage of spend.
688.32 -> It can even alert me if I'm forecasted
691.05 -> to spend more than what my budget is set for.
694.95 -> There's also Cost Anomaly Detection.
696.42 -> Cost Anomaly Detection is fantastic because
698.67 -> Cost Anomaly Detection is a machine learning model
700.8 -> that's trained on AWS cost and billing data.
703.74 -> And so that can alert me if I have an anomaly.
706.32 -> And that can be really important,
707.4 -> especially if you're just looking at monthly data, right?
709.23 -> Oh, great, we spent a $100,000 one day by accident
713.46 -> a month ago.
714.33 -> It's a little late to find that out,
715.92 -> so Cost Anomaly Detection really helps Amy figure this out.
719.335 -> Reserved instances and Savings Plans are a awesome tool for
723.9 -> FinOps and for cost optimization.
726.013 -> They're kind of table stakes at this point.
727.89 -> So that information is not necessarily in Cost Explorer,
731.97 -> so I need to go somewhere else and get this information
734.37 -> so that I can generate these reports and see
736.38 -> what have we saved? What has our coverage been?
738.9 -> What money have we left on the table?
742.5 -> - All right, so we looked at all these different tools.
746.4 -> These tools provide targeted view into the individual pieces
751.17 -> of the cost optimization puzzle.
754.02 -> But Amy wants to see all those insights in single place.
758.82 -> She wants to visualize all of that insight
761.43 -> coming from all those services from one single place.
765.6 -> And the reason for that is because
767.73 -> she wants to be able to set cost optimization KPIs
770.94 -> for her application teams.
773.55 -> She needs to understand her unit costs
775.8 -> so she can measure the impact
778.62 -> of the cost optimization efforts.
781.05 -> Now, unit costs is a way for companies to see
787.2 -> their spend in context of what it means for their business.
792.27 -> For example, if your spend is going up
796.02 -> and if your unit costs are trending down,
799.98 -> it's actually a good indication
802.95 -> that you are running with cost efficiency
806.76 -> in your environment.
807.93 -> But let's say if your spend is growing up,
810.75 -> and your unit costs are growing up,
813.18 -> that means you are leaving some
814.56 -> cost optimization opportunities on the table.
818.46 -> So unit cost also could be elusive because
821.79 -> it depends what it means to your organization.
825 -> So I'd like you to picture a three layer cake
830.13 -> with cherries on top.
832.14 -> You can add as many cherries as you want, it's your cake.
836.46 -> The bottom layer of the cake
838.92 -> constitutes your infrastructure unit costs.
842.52 -> Where you track and measure your fundamental cost
846.06 -> for your applications, for example, cost of storage per gig,
852.42 -> cost of compute per hour.
855.36 -> The middle layer forms your functional unit costs.
860.46 -> They map directly to your application teams, but their unit,
865.05 -> they're expressed in per unit terms.
867.99 -> So for example,
869.82 -> they could be the cost of processing a transaction.
872.88 -> It could be cost of uploading a file,
875.52 -> it could be cost of delivering
877.11 -> a machine learning prediction.
880.32 -> The top tier is your business unit costs.
886.02 -> They best illustrate the value of your spend in terms of
891.69 -> the value it provides to your business.
893.76 -> And just like the top layer of the cake,
896.67 -> the top layer of your unit metrics is also very unique
901.77 -> to the occasion, Father's Day, Mother's Day,
904.65 -> Valentine's Day.
906.39 -> So the top layer of your business unit metrics
909.27 -> really depends on the business you are in.
912.39 -> So now you could see that Amy wants to be able to first A,
918.54 -> derive, and B, share those insights with the stakeholders,
924.63 -> many of whom lack access to AWS Console.
928.41 -> So effectively, if you look at Amy's situation,
932.52 -> she's looking for a solution
934.47 -> that can give her a really good start, right off the bat.
938.97 -> The solution can grow with her
941.37 -> and evolve with her use cases.
943.74 -> And more importantly, she wants a solution
947.31 -> that can provide unrestricted distribution capabilities
951.6 -> so she can share those insights
953.46 -> with the teams and stakeholders.
956.01 -> So what can Amy do?
958.47 -> - This is why we built the Cloud Intelligence Dashboards.
962.73 -> So my team set about solving this problem
966.15 -> about a year and a half ago.
967.77 -> And we discovered that what we can do within AWS is,
972.51 -> offer a solution that works on AWS native services
975.48 -> on top of the underlying data sources
977.49 -> that we are talking about and deliver to Amy,
980.79 -> and everyone else who's in this position,
983.28 -> a set of dashboards in their account
985.92 -> that was built in fire, is what I'm gonna say.
990.15 -> FinOps is fire, FinOps is fire. (laughing)
993.18 -> When you've been through the pain
994.77 -> of what Amy has gone through, you do realize that,
998.595 -> it's important in how you think about this data.
1002.24 -> So when we built the Cloud Intelligence Dashboards,
1004.4 -> we realized there's actually different personas
1007.04 -> who need to tackle this.
1008.24 -> So we have six dashboards.
1010.28 -> And the six dashboards spread across these personas
1014.203 -> are our way of giving something to Amy,
1018.47 -> and giving something to the head of engineering,
1021.56 -> and giving something to Amy's boss,
1023.93 -> and giving something to the CEO, all at the same time.
1027.62 -> So the executives are gonna be really interested
1030.26 -> in the Cost Intelligence Dashboard and trends, right?
1033.62 -> These are high level summaries.
1036.38 -> All the way back onto the right,
1038.36 -> which is the Trusted Advisor
1040.01 -> and Compute Optimizer dashboards,
1041.42 -> we talked about that as a source of information.
1044.21 -> These dashboards are very FinOpsy
1047.42 -> because they are just about helping you
1051.65 -> find all of the individual resources
1053.51 -> that are over-provisioned or under provisioned
1055.37 -> and need to be right sized.
1057.02 -> And then right in the middle is the KPI dashboard
1060.05 -> and the CUDOS dashboard.
1061.43 -> The KPI dashboard is 18 cost optimization metrics
1064.13 -> that we recommend you track.
1065.75 -> We tell you where you sit with them
1067.49 -> and give you the goals to track,
1069.2 -> and what we think you would save if you met your goals.
1071.63 -> And the CUDOS dashboard is,
1073.937 -> everyone loves the CUDOS dashboard.
1076.31 -> The CUDOS dashboard
1077.143 -> is definitely our most in-depth dashboard.
1078.65 -> It gets very detailed, it allows you to drill down,
1081.11 -> start wide, drill down,
1082.79 -> and get into resource level granular information
1085.58 -> so that the engineers can take action,
1087.47 -> the product managers can take action,
1089.42 -> and everyone else can take action.
1092.81 -> - All right, so now you might be wondering
1096.11 -> how the sausage is made.
1099.53 -> So CID collects information
1102.32 -> from all the sources that we looked at.
1105.344 -> AWS cost and usage report,
1107.54 -> the most comprehensive source of your billing information.
1110.93 -> We collect data from AWS organizations,
1113.39 -> which is the metadata about your accounts,
1116.15 -> AWS Budgets, AWS Trust Advisor, as well as Compute Optimizer
1122.61 -> where we gain insights from over or underutilized instances.
1128.24 -> And then if you look at the bottom, the CID framework,
1131.96 -> that's the core artery through which all these insights
1135.08 -> are made available to QuickSight,
1136.97 -> and in turn to Cloud Intelligence Dashboards.
1140.12 -> In more technical terms, we are giving you a fully automated
1144.59 -> serverless analytics pipeline that is forming
1148.58 -> a cost data lake for you to turbocharge your FinOps journey.
1156.56 -> Now let's take some of these dashboards.
1158.87 -> - Should we show them? - Yeah, who wants to see them?
1162.17 -> - Nobody, everybody wants to see them.
1164.086 -> (laughing)
1172.58 -> There we go.
1174.14 -> HDMI 1, everybody, just in case you were wondering.
1177.08 -> Okay, so what you're looking at right now
1180.71 -> is the Compute Optimizer dashboard.
1182.66 -> So we usually recommend our customers
1184.61 -> start with this dashboard,
1185.9 -> because we can have an argument about this
1189.17 -> and we probably should, after we're done,
1192.08 -> about what's the right order in tackling FinOps
1196.16 -> with all of the options that you have.
1197.87 -> But right sizing is a really good thing to do
1200.87 -> pretty early on in the process.
1202.94 -> So the Compute Optimizer Dashboard
1204.56 -> is here to help you reach right size.
1206.96 -> Now this is all the information
1208.58 -> from all your payer accounts, all your organizations,
1210.65 -> all in one place.
1211.483 -> It doesn't matter what region they're in,
1212.9 -> it doesn't matter what the account is.
1215.3 -> I've got $520,000 worth of, we could say waste,
1220.43 -> we could say potential savings.
1221.69 -> It's a philosophical difference, you decide,
1224.178 -> that I need to tackle.
1226.25 -> I know that down here, this month over month graph
1229.46 -> is actually giving me something very interesting
1231.89 -> that I don't get anywhere else, which is context.
1234.98 -> So if I told you that you had $258,000
1239.6 -> worth of potential savings, that might sound like a lot,
1243.23 -> it might sound like not a lot.
1245.24 -> Without knowing what was it last month,
1247.28 -> it's hard to tell if you're headed in the right direction.
1249.41 -> So throughout these dashboards,
1251.24 -> you're gonna see month over month graphs for that reason,
1253.73 -> because I can't just give you the state of the state today
1257.69 -> and the facts today without giving you the context.
1260.39 -> Now this is a very high level summary
1262.64 -> of all of the right sizing we need to do.
1264.86 -> We've got about a quarter of a million dollars
1266.6 -> worth of right sizing to do with EC2,
1268.82 -> and some right sizing to do with other sources.
1271.64 -> So I'm just gonna show you the EC2 tab today
1274.52 -> 'cause we, you know, for time.
1276.32 -> In the EC2 tab,
1277.55 -> I have the option of looking at all of this information
1280.1 -> by business unit, which is something I really like to do.
1282.95 -> And that's because you might have one account
1286.46 -> for business units, some customers do that,
1288.26 -> but a lot of customers don't.
1289.16 -> They kind of grew organically on AWS
1291.077 -> and now they've got a thousand accounts
1293.09 -> and, you know, 15 business units, right?
1295.7 -> So that grouping can easily be imported into this dashboard,
1299.54 -> into QuickSight,
1300.373 -> in fact, it can be imported into all the dashboards you see.
1303.11 -> And I can sort of get a better picture of
1305.9 -> how my business units are doing in terms of their
1308.114 -> opportunity for right sizing and optimization.
1310.76 -> Now again, this month over month graph tells me
1313.55 -> I am headed in the wrong direction because
1315.86 -> my right sizing opportunities are increasing,
1318.05 -> which is not what we want.
1320.15 -> When I look at this graph here by business unit,
1322.22 -> I can see that the business unit, Huff Schwartz and Young,
1324.56 -> which is made up, needs the most attention, right?
1328.46 -> We all have limited resources.
1330.53 -> I've not heard of a customer who's got
1332.603 -> 1000 FinOps people working for them quite yet, right?
1335.75 -> I don't think we're there yet, right? These teams are small.
1338.42 -> And so sometimes we need to prioritize.
1341.66 -> So this helps us prioritize where we need to take action.
1344.45 -> Now all these visuals are gonna help us kind of
1346.37 -> slice and dice the recommendations
1347.99 -> and understand what's the highest risk?
1350.21 -> What's the highest amount of potential savings?
1352.22 -> And then the engineers
1354.05 -> should have access to this dashboard too,
1355.82 -> because here is literally every single instance
1358.49 -> where they are and what the right sizing recommendation is.
1361.22 -> So if I click on this one for example,
1363.08 -> it's an R54X large, it's over-provisioned.
1366.454 -> If I were to move it to a T3 large,
1368.84 -> which is the recommendation, I'd save quite a lot of money,
1371.8 -> $1,200 to be exact per month.
1374.027 -> And I can make, I can check these statistics here
1378.29 -> about the percentage of memory in CPU that's being used,
1381.08 -> in my case it's under 9% of CPU and under 4% of memory,
1384.5 -> so I'm barely using this instance.
1386.42 -> And this is over a period of time
1388.4 -> to make sure that we're taking those burst workloads
1390.95 -> into a consideration as well.
1392.15 -> So we wanna make sure you have this information
1394.226 -> before you actually go and change
1395.9 -> and right size this instance.
1397.76 -> So the Compute Optimizer dashboard is great for those
1400.923 -> right sizing recommendations from EC2, Autoscale,
1404.12 -> Lambda, and EDS.
1406.13 -> The Trusted Advisor dashboard
1408.71 -> is another really great resource
1410.99 -> for this kind of information.
1412.1 -> Now, Trusted Advisor has a lot of other non-cost related
1415.19 -> checks that it does.
1416.66 -> Again, month over month graphs.
1418.7 -> Why? Because now I have some context,
1420.89 -> I can see that my teams are getting closer to green overall,
1424.43 -> so I don't have to yell at anybody.
1426.59 -> If we go to the cost optimization tab,
1429.05 -> we're gonna get even more right sizing recommendations.
1432.532 -> If I scroll down here some of the key ones
1435.14 -> that I'll show you and there is some overlap by the way,
1437.63 -> between Trusted Advisor and Compute Optimizers.
1439.64 -> So this is not a perfect world,
1442.1 -> and there's always room for improvement,
1443.66 -> but it's important for me to have all the data first.
1446.09 -> Now I know what my idle RDSDB instances are, right?
1449.6 -> So I can see what that is,
1451.04 -> I can get a list of the actual instances where they are,
1453.53 -> what the resource idea is,
1455.06 -> and how much I could potentially save, right?
1457.43 -> So now I know whose phone number to call and say, excuse me,
1461.153 -> we're spending a lot of money on these RDS instances,
1463.64 -> what's going on?
1465.17 -> There's EC2 instances is covered, and then we have some,
1468.56 -> other rightsizing recommendations that
1471.23 -> when you explore these dashboards,
1473.06 -> which you will all be able to do,
1474.68 -> this link is publicly available
1476.24 -> and I'll make sure we share it at the end
1478.43 -> so that you can all see what these dashboards look like
1480.83 -> and interact with it.
1482.93 -> Now I'm gonna jump to the KPI dashboard, because I think
1486.05 -> that this is a really good next step for Amy,
1488.87 -> and for folks who are in the FinOps role.
1492.2 -> The KPI dashboard, as I mentioned earlier,
1495.2 -> has 18 cost optimization KPIs that we recommend you track.
1499.7 -> And again, remember FinOps is fire, we were in the fire.
1503.12 -> These are the ones that customers have told us
1505.04 -> are really important to them.
1506.15 -> So we didn't just pull these out of thin air.
1508.49 -> And these the low hanging fruit,
1514.07 -> the way that you can prioritize this work
1515.87 -> given limited resources.
1517.07 -> So things like what percentage of your spend
1519.47 -> is on Previous Generation instances?
1521.42 -> What percentage of your spend is on Spot, on Graviton,
1524.81 -> on on-demand?
1525.8 -> What percentage of your spend is on old snapshots,
1528.32 -> GP2 volume types? Those kinds of things.
1531.98 -> This is gonna give me a really good insight
1533.6 -> into where I sit in my green, what's my organization doing?
1536.84 -> Look, I've got 1% of my spend on Graviton,
1539.21 -> 40% of my spend on Spot,
1540.62 -> that's why it's green, 'cause it's good.
1542.24 -> And I've got 4% spent on Previous Generation instances.
1544.639 -> So I'm doing pretty well here,
1546.53 -> but there's a lot more room for improvement.
1548.21 -> In fact, I can see what my potential savings would be
1551.18 -> if I met these goals.
1552.5 -> Now the goals are set by us in the dashboard,
1555.59 -> and they're pretty generic.
1556.7 -> So for example,
1557.72 -> we recommend that less than 10% of your spend on EC2 be on
1561.44 -> Previous Generation instances.
1562.73 -> We recommend that less than 30%
1565.58 -> of your spend be on on-demand.
1567.5 -> Change these goals, right?
1570.522 -> This is meant for you to interact with,
1572.87 -> this is meant for you to come in and say, okay,
1574.64 -> well that's a good starting place,
1576.89 -> but man we want to be really aggressive with Graviton
1578.767 -> 'cause we just attended an awesome re:Invent session
1581.24 -> and we learned about Graviton7,
1583.16 -> and it's gonna save us a whole lot of money, right?
1585.47 -> So this lets me come in here as an organization
1588.2 -> and decide what these need to be,
1589.61 -> and you can be at a team level,
1590.87 -> it can be at an application level.
1592.43 -> And when I come back,
1593.33 -> I can see how I track against these goals,
1595.57 -> and then I can see the potential savings again adjusted.
1599.242 -> (coughing) Excuse me.
1601.52 -> In these other tabs, I'm gonna find more details.
1603.8 -> Like in the ECT2 tab it's gonna break down
1606.56 -> what my potential savings would be with AMD
1609.167 -> and Graviton instances.
1611.27 -> And I can see what my past savings has been,
1613.34 -> because I tried them.
1614.66 -> I implemented them in one of my accounts,
1616.88 -> and now I can see what actually
1618.32 -> my estimated potential savings had been in the past,
1622.1 -> which helps us inform what they could be in the future.
1624.95 -> But look, I've got $88,000 worth of savings
1627.92 -> by switching to Graviton, $33,000 by switching to AMD.
1631.4 -> These have different costs to do.
1633.92 -> So AMD instances are not ARM 64 instances like Graviton are,
1639.59 -> so I might not need to refactor my code
1642.02 -> as I would for Graviton, but the savings is different.
1644.63 -> Now I have the information to know which decision to make,
1647.84 -> and I also know which to prioritize,
1650.21 -> because now I have all the instances.
1651.89 -> This is every single instance that we recommend
1654.653 -> you migrate to, either AMD or Graviton
1657.35 -> based on the instance type.
1659 -> We have similar tabs for EBS and S3.
1662.06 -> They're gonna help you find old snapshots, stale buckets,
1664.73 -> your savings from infrequent access and glacier,
1667.649 -> and your savings from migrating from GP2 to GP3.
1671.69 -> Now it's time to come to the CUDOS dashboard.
1675.32 -> So CUDOS is gonna really help me dive deep
1679.22 -> into my cost and usage information.
1681.56 -> There's some high level stuff here that's gonna help me like
1683.87 -> what was my invoice spend month over month?
1685.97 -> How does it break out?
1687.14 -> How much of it was refunds, taxes, credits?
1689.45 -> How does it break out by tag?
1690.98 -> How do I charge back to my organizations?
1693.14 -> I can do that with my amortized costs,
1695.03 -> I can do that with my invoice spend.
1697.04 -> All of that is here, and I can customize this.
1700.52 -> Anything you see here can be customized.
1702.17 -> I can add alerts, I can add forecasts.
1704.349 -> One of the graphs that I really particularly like in CUDOS
1709.58 -> is the month over month trends.
1711.23 -> Because not only is this gonna tell me
1713.36 -> what's shaken and baking, right?
1715.13 -> What's changing month over month
1717.2 -> based on the amount of change,
1718.88 -> but I can also see a list of all of the services here
1721.67 -> and what their month over month changes.
1723.95 -> And I'm seeing a lot of green arrows pointing down,
1726.56 -> meaning I'm spending less on these services
1728.27 -> and sometimes they're up.
1729.98 -> And if I see something that's up on a given month,
1732.62 -> perhaps the most recent month,
1734.21 -> I can click on it and say, okay, what's causing that change,
1739.501 -> and from which service, and which account is responsible?
1744.35 -> So I'm just gonna expand this so you can kind of
1746.06 -> see it a little bit better, right?
1747.59 -> So simple storage service here,
1750.65 -> I now see which accounts are spending
1752.36 -> on simple storage service, S3.
1754.28 -> I can scroll down, everything else is now filtered
1756.17 -> just to that application.
1757.61 -> I can see which accounts are responsible,
1759.29 -> this account is clearly spending the most money on S3,
1763.04 -> it's just way more than everybody else.
1765.26 -> It's all API requests, which is interesting,
1767.78 -> it's not on the storage.
1769.16 -> So it's not like they put giant files in there
1770.93 -> and they're just hanging around, they're doing something.
1773.58 -> So now I have some more information,
1775.82 -> I can drill down further
1777.38 -> and learn what the API requests are.
1779.51 -> So I see that they're GET objects,
1781.31 -> and they kind of are increasing month over month.
1783.44 -> So now in a few seconds I was able to drill down
1785.93 -> on a trend that I caught.
1787.55 -> Now you can switch this to week over week trends,
1789.65 -> whatever floats your boat.
1791.21 -> Another tab that's really popular here is data transfer.
1794.21 -> So the data transfer summary tab
1796.88 -> is we break out all of the data transfer costs that
1799.97 -> oftentimes get bundled with other services
1801.71 -> so that you now know what's going on.
1804.05 -> Now note,
1805.237 -> in CUDOS you're gonna see a lot of recommendations.
1807.62 -> These are static recommendations.
1809.36 -> These are cost optimization recommendations.
1813.56 -> And please read them, please go through them,
1815.54 -> and tie them to the data that you're seeing in CUDOS.
1818.06 -> So Amy has all of this insight
1820.22 -> from all of the terms and the SAAS
1821.66 -> and all the engineers all around the world
1823.13 -> who've been doing this for years and figured out,
1825.519 -> what's gonna have the biggest impact?
1827.644 -> We can help you find that.
1829.94 -> And so for example, with data transfer,
1833.39 -> I can find daily spikes like this,
1835.04 -> I see there was a big spike on that day.
1836.81 -> I can drill down deeper into the services
1838.97 -> that are responsible for the data transfer costs
1842.24 -> and I can see the breakout by month, by day, by account,
1846.35 -> by region, and then again now by resources.
1849.38 -> So this is telling me what my top 10
1850.85 -> most expensive data transfer resources are, right?
1854.06 -> That's really useful information for me to know,
1855.8 -> because now I can figure out, well maybe,
1857.78 -> I need to put everything into a single AZ.
1861.14 -> Maybe I need to switch to some other managed service
1863.9 -> for data transfer that reduces the costs
1865.835 -> like moving from Kafka to the managed service
1869.75 -> that Amazon has.
1871.61 -> All of these kinds of decisions I can make
1873.68 -> based on the data that now Amy has.
1877.13 -> So I didn't cover all of the dashboards today,
1879.784 -> just because of time.
1882.29 -> So we're gonna move on,
1883.64 -> but if you have any more questions about these dashboards,
1886.55 -> then just please let me know after the session here.
1897.32 -> There we go.
1898.153 -> - All right, so all that you saw was
1902.42 -> built on top of QuickSight.
1906.65 -> Now scaling financial awareness is the first step
1911.39 -> to creating a cost efficient culture
1913.49 -> and driving efficiencies within your organization.
1916.1 -> So with Cloud Intelligence dashboards powered by QuickSight,
1919.43 -> a multitude of ways of doing that.
1921.11 -> First is you can share these dashboards
1924.52 -> with everyone in your organization.
1928.1 -> And QuickSight provides enterprise grade control
1930.89 -> to administer the permissions.
1932.6 -> You can also take advantage of role level security
1936.44 -> and column level security to control who gets to see what.
1941.42 -> For example, you may wanna scope these dashboards
1944.45 -> to certain AWS accounts, based on the user who's logged in.
1949.4 -> You can easily express that use case.
1952.76 -> Third, you can deliver these insights to your stakeholders
1958.52 -> directly in their applications they use
1960.89 -> on a day-to-day basis.
1962.93 -> QuickSight provides a rich set of APIs to embed
1966.47 -> these dashboards and visuals into web applications,
1969.89 -> internal VQs, Runbooks.
1972.65 -> And we also provide you options
1974.33 -> to customize the look and feel
1976.37 -> so they're indistinguishable from your host application.
1980.9 -> You can also configure QuickSight queue
1984.77 -> with Cloud Intelligence dashboards
1987.02 -> to enable natural language interface
1992.96 -> to be able to express questions in plain English.
1997.16 -> You could potentially ask
1998.27 -> what was my storage spend last month?
2001.15 -> And have a visual return to you
2004.33 -> without having to pre-bake those insights
2006.01 -> into the dashboard.
2007.48 -> You can also ask why questions that we announced.
2010.39 -> Which is why was my spend for storage up last month?
2014.89 -> And QuickSight will run correlations in background
2017.05 -> and will return you the probable causes.
2020.26 -> So you can see with the power of these visuals
2023.86 -> and the distribution capabilities in QuickSight,
2026.77 -> you can truly create end to end financial awareness
2029.62 -> within the organization.
2031.93 -> So now coming back to Amy,
2034.48 -> she learns about all these capabilities
2037.72 -> and she works with her AWS admin,
2041.23 -> and runs through the setup,
2044.17 -> and launches these dashboards in her account.
2046.93 -> And now she's able to quickly report on
2050.5 -> the monthly spend trends,
2052.36 -> she's able to get to the resource level details
2054.79 -> so she can understand what was the cost
2056.59 -> of a Lambda function?
2057.52 -> What was the cost of a DynamicDB table?
2059.95 -> So she's able to get both the bird eye view,
2063.31 -> as well as the where I'm I view from these dashboards.
2067.33 -> She's able to see all this right sizing recommendations
2070.09 -> from a single place and she can show more importantly
2073.75 -> to her CFO how her spend is trending
2077.11 -> and how that's helping her meet her objectives.
2082.369 -> Now, by utilizing Cloud Intelligence dashboards,
2087.16 -> both Amy and Martha have a strategy in place,
2093.19 -> where Martha is able to get spend updates
2097.03 -> personalized for her in the format she prefers
2100.93 -> at the frequency she wants.
2103.609 -> Amy feels empowered, because she can lead with data
2107.98 -> and insight with her engineering teams
2110.11 -> to drive cost optimization efforts.
2113.59 -> And with that,
2114.46 -> now they have a continuous cost optimization strategy
2117.28 -> that allows them to see the spend
2119.98 -> so they can maximize their investments with AWS
2123.4 -> and reach their goals faster.
2126.34 -> So that was Amy.
2129.55 -> Now let's hear from Mike,
2131.763 -> who's the FinOps practitioner from Dolby.
2135.25 -> - Real customer ladies and gentlemen.
2138.062 -> (audience clapping)
2148.506 -> - Hi everybody, I'm Mike Graff.
2149.44 -> I'm the lead infrastructure architect
2150.79 -> for Dolby Laboratories.
2152.74 -> So for those that may not be familiar with Dolby,
2155.38 -> we are an entertainment technology company
2157 -> that was founded 57 years ago
2158.56 -> by a pioneering inventor Ray Dolby.
2161.41 -> Now, when Ray founded Dolby in 1965,
2164.65 -> movies and television featured only one channel of sound.
2168.16 -> And record producers were limited
2169.39 -> to a handful of audio tracks.
2172.472 -> Much of what has happened since then
2175.157 -> to improve the sound of entertainment
2177.01 -> can be traced back to Ray.
2178.51 -> Not just to technical innovations,
2180.296 -> but the impact they had on artists.
2183.79 -> Ever since our founding, our vision has been to
2187.51 -> revolutionize the science of sight and sound
2189.52 -> through our innovative research and technologies.
2192.28 -> We empower creatives to elevate their stories,
2195.43 -> and we offer fans incredible experiences.
2199.09 -> Our history of entertainment innovation began
2201.46 -> in the 60s with Dolby noise reduction.
2203.62 -> You all probably remember the noise reduction button
2205.51 -> if you're old enough to know what a cassette tape is.
2208.698 -> In the 80s we had Dolby surround,
2211.18 -> and now the innovation continues today
2213.16 -> with our groundbreaking technologies
2214.51 -> like Dolby Atmos for Sound, and Dolby Vision for imaging.
2220.12 -> In 2020, we brought our innovation to the Cloud,
2222.79 -> with Dolby.io, which is our developer platform.
2226.27 -> It's a collection of APIs and SDKs
2228.13 -> to help support immersive, interactive,
2230.53 -> and social audio and video experiences.
2234.19 -> This includes things like real time streaming,
2236.53 -> voice and video chat, media processing, and music mastering.
2241.51 -> We have customers that use these services
2243.34 -> for a wide variety of use cases
2244.78 -> such as AR and VR experiences, webinars,
2248.17 -> live and virtual events, sports and betting,
2251.32 -> remote production, podcasting, and more.
2255.88 -> You can scan the QR code here on the slide
2258.34 -> for more information about Dolby.io
2259.9 -> and experience some demos for yourself
2261.37 -> and even sign up for an account.
2264.13 -> Now I've been lucky enough to participate in Dolby's journey
2266.83 -> for the past 22 years.
2269.08 -> And today I'm gonna share a little bit with you about our
2271.63 -> Cloud FinOps journey.
2275.35 -> So we began using public cloud back in 2015,
2278.74 -> and at that time we were just starting to crawl
2281.05 -> with our cloud financials.
2283.42 -> Back then our finance department
2285.01 -> was not a fan of chargeback.
2286.3 -> I'm sure there are folks in the audience
2287.74 -> who can relate to that.
2289.78 -> So as a result,
2290.613 -> we outsourced the whole thing to a third party reseller.
2293.41 -> We worked with a reseller, we built a complicated system
2295.87 -> where we consolidated all of our spend
2298.09 -> with an account managed by the reseller.
2300.372 -> Each business team would have to issue a separate PO
2303.37 -> to the reseller for their spend.
2305.38 -> And then the reseller was using a commercial
2308.65 -> cloud cost management tool to slice and dice our bill,
2311.8 -> generate 70+ invoices each month to Dolby,
2315.04 -> and charging us just 3% of our spend for that service.
2318.387 -> (audience laughing)
2321.04 -> So fast forward to the second half of 2021,
2324.839 -> and at that point we realized the amount of money
2326.29 -> we were spending on AWS
2328.21 -> combined with our increased maturity
2329.68 -> around managing public cloud
2331.42 -> meant that it was time to start walking on our own
2333.61 -> and bring the billing in house.
2336.25 -> So in April of last year, I wrapped up a six month,
2338.83 -> actually April of this year,
2340 -> I wrapped up a six month effort to identify and implement
2342.7 -> new internal tools to replace the commercial tool,
2346.3 -> work with our finance group to design and implement
2348.49 -> a cost allocation scheme for our cloud spend,
2351.58 -> and migrate away from the third party reseller.
2355.21 -> The result has been a lot less paperwork
2357.28 -> in the form of purchase orders, invoices, payments.
2360.79 -> More transparency on AWS costs for our account holders,
2364.21 -> and stronger FinOps muscles for our team.
2370.33 -> So when I started the project
2371.41 -> to move our cloud billing in-house,
2372.79 -> one of the first things I did was perform a gap analysis
2375.04 -> to figure out what were the capabilities
2377.14 -> we needed to replace from the commercial tool
2378.91 -> that was gonna be going away?
2381.34 -> And very early in the process,
2383.14 -> my AWS TAM informed me about the existence
2385.24 -> of these Cloud Intelligence dashboards.
2387.97 -> The solution was quick and easy to set up,
2389.62 -> following the well architected labs
2391.24 -> that walk you through the process.
2393.52 -> And once deployed, we were able to start customizing
2395.71 -> the dashboards to fulfill a lot of the requirements
2397.57 -> that we identified in the gap analysis.
2401.71 -> And then combined with another AWS benley feature called
2404.26 -> Cost Categories, we were able to deliver a complete,
2407.662 -> cost allocation reporting system
2409.81 -> that works for both our cloud teams
2411.67 -> as well as our good friends in the finance department.
2415.72 -> And best of all, it's an extremely cost effective solution.
2419.44 -> Our production deployment includes
2421.3 -> data for over 160 AWS accounts.
2424.48 -> We share these dashboards with more than 50 users
2426.58 -> across six different business groups.
2428.86 -> Our monthly cost for this feature is $150 a month,
2433.99 -> which is significant cost savings
2435.91 -> over that previously mentioned 3% of our spend.
2443.11 -> So we focused our initial deployment on the CUDOS dashboard
2445.93 -> that Aaron was just talking about.
2448 -> I felt like it gave us the broadest compliment
2449.86 -> of those top level spend visualizations
2452.65 -> as well as the deep dives into specific services
2455.201 -> that mattered to our stakeholders.
2459.34 -> In addition, it easily allowed our users to self-service
2462.13 -> the data they needed
2462.97 -> for the specific accounts they were interested in
2464.92 -> and filter out the rest.
2467.65 -> And then CUDOS also gave us a rich set of visualizations
2470.77 -> that we could then customize to our own needs.
2473.89 -> At this point, we've even gone so far as creating
2475.81 -> custom versions of the CUDOS dashboard
2477.85 -> for individual business teams
2479.59 -> and just only remove the tabs they're not interested in
2482.44 -> and only have the tabs that they care about.
2489.16 -> In addition, we really liked the ability
2491.17 -> to bring in all our custom defined cost allocation tags,
2494.86 -> our cost categories that I mentioned before.
2497.14 -> We were able to bring those all into the dashboard
2498.7 -> and allow us to build custom reports
2500.65 -> based on spending by specific tags,
2502.42 -> which is something that a lot of our teams
2503.86 -> really care about.
2506.71 -> In addition, customized views based on the cost categories
2510.85 -> has allowed us to look at adding some gamification
2513.19 -> to our cost optimization efforts,
2516.19 -> by showing business groups how they're performing
2518.23 -> against other business groups.
2519.46 -> So like on this graph on the bottom left,
2522.94 -> I've got a custom widget we built out
2524.56 -> that shows how well each cost center is doing
2526.57 -> on adding Spot instances and Savings Plans
2529.75 -> into their compute mix.
2531.34 -> And on the lower right, I'm showing each cost center
2534.91 -> how much money they're saving
2536.23 -> by using those savings instruments.
2543.88 -> One of the most valuable views we've gotten from CUDOS
2546.04 -> is the ability to drill on to data transfer costs,
2548.71 -> this is something Aaron was talking about.
2550.84 -> I'm sure you can all relate
2551.77 -> that this is something that can be pretty opaque
2553.72 -> on the AWS bill.
2555.58 -> And it's often a source of much consternation
2558.13 -> among your account owners.
2560.74 -> The visualizations available on the dashboard
2562.45 -> help to remove a lot of the mystery around
2564.13 -> where your data transfer costs are coming from.
2567.743 -> And in addition, they can help identify potential
2570.58 -> architectural issues in your environment.
2573.37 -> A real world example for Dolby
2575.32 -> was when we were able to leverage CUDOS
2577 -> to help us identify a misconfiguration
2578.95 -> in one of our Kubernetes clusters,
2580.87 -> that was we were creating an excessive amount
2582.49 -> of inter AZ traffic.
2585.34 -> The data transfer costs by type visualization
2587.95 -> you see on the right helped us to identify and eliminate
2590.89 -> about $7,000 a month in inter AZ costs
2594.043 -> in a single group of accounts.
2597.891 -> In addition, we've recently used
2599.86 -> that same data transfer cost to have to identify
2602.29 -> accounts that are spending too much money on Nat Gateway.
2607.03 -> The solution was a quick and easy 30 second fix
2609.7 -> to implement the S3 endpoint,
2611.74 -> and eliminated over $8,000 a month in internet charges
2616.39 -> for a single account.
2622.3 -> So here I've got a quote
2623.2 -> from one of my business stakeholders,
2624.52 -> where he highlights the value
2625.57 -> that the Cloud Intelligence dashboards
2626.92 -> has provided to his team.
2629.14 -> For them, the dashboards have really helped them
2631.33 -> to flash out the unit cost that Rohit was talking about.
2635.32 -> They're building a new service in the cloud
2636.91 -> and they wanna understand what the cost
2638.14 -> of that service is gonna be, and how it's gonna scale.
2640.24 -> And these dashboards are really helping them with that.
2647.23 -> Now I wanna talk about a couple of the enhancements
2649.3 -> that I put in place
2650.32 -> to make the Cloud Intelligence dashboards more usable
2652.45 -> in our environment.
2654.46 -> First was implementation of AWS SSO access for QuickSight.
2659.521 -> And because we have these QuickSight dashboards
2662.23 -> running in our payer account,
2664.06 -> I really didn't want to give all my users who needed access
2666.523 -> login capabilities to my payer account.
2670.15 -> By implementing SSO for the service,
2671.68 -> we're able to just grant users access to QuickSight directly
2674.2 -> without granting any additional Console access
2676.39 -> to the payer account.
2678.61 -> Account owners are just shown another button
2680.59 -> on their SSO portal that they're already using
2682.78 -> to log into the AWS Console.
2686.65 -> Now users are provisioned
2687.82 -> with QuickSight user accounts automatically
2689.65 -> the first time they log in with AWS SSO.
2692.74 -> And once they do that,
2693.64 -> we can then place them into QuickSight groups,
2695.92 -> which enables the second enhancement we put in place,
2698.11 -> which is called role level security.
2700.24 -> Rohit also had also talked about that earlier.
2702.82 -> So now by default, when you set up these dashboards,
2704.71 -> when a user logs in, they have ability to see cost data
2707.59 -> for all the accounts in your environment.
2710.62 -> Now if you have a large environment
2712.15 -> with hundreds or thousands of AWS accounts,
2715.12 -> that might be less than desirable.
2718.656 -> For the first part for the user,
2720.61 -> it might be hard for them to figure out
2722.26 -> which accounts do I care about, right?
2723.7 -> I don't know which ones are mine.
2725.98 -> And as an admin,
2726.94 -> you may have a desire to limit what they can see.
2730.39 -> So with role level security,
2731.59 -> we can create a rule set that maps QuickSight accounts,
2734.17 -> sorry, QuickSight users or groups to AWS accounts.
2737.5 -> And then you apply this mapping to the QuickSight data sets
2741.13 -> that are involved in the Cloud Intelligence dashboards.
2744.28 -> And the result is that the users are only able to view
2746.41 -> cost data for the accounts
2747.67 -> which they've been driven access to.
2754.75 -> So if you're interested in learning more about
2756.13 -> how to implement these enhancements in particular,
2757.96 -> I'll make a shameless plug for my own blog.
2761.47 -> You can go into great detail on how to set these features up
2764.2 -> as well as how we're leveraging cost categories
2766.99 -> in conjunction with the Cloud Intelligence dashboards.
2770.47 -> You just scan the QR codes here
2771.94 -> and they'll take you to this respective blog post
2773.8 -> on these topics.
2779.26 -> So what's next for Dolby?
2781.54 -> With the initial success of CUDOS
2783.76 -> in our Dolby's environment, we're now rolling out
2785.59 -> some additional dashboards that Aaron talked about
2788.05 -> to help our teams identify and implement
2789.58 -> savings opportunities in their accounts.
2792.7 -> One of the dashboards we really like is the KPI dashboard.
2795.76 -> It's a great solution that allows teams
2797.8 -> to set personalized KPIs like Aaron was showing you
2801.58 -> for various cost optimization categories, like,
2804.061 -> Spot Instances in use, GP3 versus GP2,
2807.34 -> or how many snapshots do you have more than a year old.
2812.62 -> Once you set these personalized KPIs,
2815.11 -> they can then view their progress
2816.49 -> towards each goal at a glance,
2818.35 -> as well as get a historical view
2820.09 -> to see how things have hopefully improved.
2825.16 -> We're also starting to roll out
2826.24 -> the Compute Optimizer dashboard.
2828.94 -> Compute Optimizer itself is a great built-in service
2831.34 -> within an AWS console, but getting reports outta that tool
2833.77 -> can be a little bit cumbersome.
2835.45 -> I don't know if anyone can relate to that, but it's not fun.
2839.14 -> With this dashboard, it makes it a lot easier
2841.36 -> to see specific areas they can be optimizing
2843.31 -> their compute spend, as well as get specific
2846.67 -> instance right sizing recommendations,
2848.41 -> like Aaron was showing you,
2849.46 -> including how much money you're gonna be saving.
2852.88 -> Now, the initial teams that I've shared this dashboard with
2855.37 -> have absolutely loved it and they can see they can,
2857.62 -> they believe they can immediately start getting value
2859.57 -> out of it.
2860.98 -> And this is particularly important
2862.39 -> in the current economic environment,
2863.77 -> where everyone is particularly focused on
2865.93 -> cloud cost optimization.
2872.95 -> So in closing, I just wanna sum up the business benefits
2875.98 -> we've seen from implementing these dashboards.
2879.4 -> The first is it's a self-service portal.
2881.95 -> One stop for spend reporting, utilization metrics,
2885.25 -> and those wonderful cost optimization recommendations.
2890.35 -> Secondly, the tool gives the cloud teams
2893.32 -> the ability to gain insights across all their accounts
2896.07 -> in a single view.
2897.37 -> So right on having to log into Cost Explorer
2899.95 -> and each end account individually,
2901.21 -> they can see it all in a unified view
2903.4 -> and get those cross account insights.
2907.27 -> And then next, it allows my team
2910.21 -> to report on progress towards cost optimization
2912.46 -> across our entire organization, and track that progress.
2917.68 -> And last but not least, who doesn't like saving money?
2921.542 -> As I mentioned before,
2923.32 -> this is like the cost of a couple pizzas a month,
2926.17 -> compared to some of these commercial
2927.76 -> cloud cost management tools which are
2930.7 -> very capable but also much more expensive.
2938.65 -> So I hope you found my story interesting,
2941.77 -> and then you found something
2942.603 -> you can take back to your organization.
2944.74 -> I wanted just leave you with a quote here from our founder,
2946.84 -> Ray Dolby, which kind of sums up the inventor's spirit.
2951.49 -> I hope that you'll take this as an inspiration
2953.35 -> to go out and build something on your own.
2957.46 -> Thank you very much.
2959.409 -> (audience clapping)
2963.49 -> - Stay up here with us.
2965.95 -> All right, so what you heard
2968.17 -> sounds like something you needed yesterday.
2971.14 -> You can launch Cloud Intelligence dashboards today
2973.45 -> from well architected Labs.
2975.43 -> Once you set up those labs,
2977.77 -> the QuickSight links would be available for you.
2980.41 -> Once you open up those dashboards,
2982.06 -> you're gonna see hundreds of predict visuals that provide
2985.51 -> insights into the most common indicators
2987.76 -> customers are looking at.
2989.29 -> You can customize them, you can personalize them,
2993.16 -> and you can share with them with your stakeholders.
2996.73 -> And you can also bring in additional data
2998.44 -> to build even more complete picture of your cost.
3004.44 -> Thank you all.
3006.566 -> (audience clapping)

Source: https://www.youtube.com/watch?v=BNYhk_wm4no