Kick-Start Intelligent Transformation with SAP and Google Cloud Platform (Cloud Next '19)
![Kick-Start Intelligent Transformation with SAP and Google Cloud Platform (Cloud Next '19) Kick-Start Intelligent Transformation with SAP and Google Cloud Platform (Cloud Next '19)](/pic/kick-start_intelligent_transformation_with_sap_and_google_cloud_platform_cloud_next_19_/ekZUbmFlOHJrWkU_.jpeg)
What. An exciting start here, I'm not sure about you guys but I'm. Really excited about. After. Yesterday's, keynotes, to, hear about how, much focus Google, is giving on ASAP. And all. The exciting news, and innovation, that is coming along. With us right so. Thinking. About 6. And 12 terabytes of, Hana machine, talking. About anthos, and. Live. Migration all of those are really exciting, news and, I'm. Sure I was, here last year also we talked about sapa little bit but, this year the excitement, has grown a lot, so. We'll talk a little bit about how. We see, the reaction, in the market what, our clients are doing and what. Google and a say P and Deloitte are doing together but. Before we start maybe. We start a little bit of introduction. So, I'll start myself Pavan Shrivastav, I focus, on building, our sa pn cloud practice, i've. Been doing a say p work for last 20. Years and in, last 5 years focused, around cloud so, we'll start with ray murphy first survey wasn't morning everyone my. Name is Ray Murphy I work at Google. Right. Now I'm, an engineering director in, our corporate engineering, so. We build the products that make Googlers. More efficient, in their job, we're. Focused on financial, systems and accounting, systems right now I've, been doing. Financial. Systems. And accounting, work for about 25. Years. Thank. You Ray pravin, morning. Everyone how Praveen miler I'm, with Deloitte, I'm gonna say P cloud architect, been doing sa P for last 19. Years designing. Architecting. And deploying solutions, on on premise for, the last five, years has been focused around helping. Migrate. All this ap workloads on to the cloud design deploy, solutions on cloud as well great. Thanks, so so. I think I think we want to start talking a little bit about what's. Happening, in this world around, us ap as you. Heard in yesterday's keynote as well that, sa P has come off the new product line as for, whereas. A P would like to see all of their customers. Who are currently on ECC, to, slowly migrate, and move to as far as for Hana by 2025. At, the same time cloud. Is also very eminent topic, in everybody's. Organization. And they, are trying to look at ways of, embracing. Cloud adapting, cloud or using, all the innovations, on cloud at the, same time and, and. We are seeing those two things coming together at, the same time and looking at how organisations are leveraging. The benefit of those I. Know Google themselves are embarking. Upon a transformation, journey and sort. Of eating your own dog food so, we would like to hear from. Ray a little bit about, what's. Your, business challenges how, you are. Looking. To adopt these technologies, in your organization, but, I'm. Sure Google is big organization. You. Are one of the top in your own data situations, your own volume and transaction, is much larger, and bigger so, talk to us a little bit about what's, your situation here so, we. Started. You. Know, you. Know many years ago was a single, business unit, single. Business model, single revenue stream and in, the past you. Know 20 years became a very complicated company, we. Grew fast, we grew. Vertically. Horizontally. But. Our systems really didn't grow in our. Treasury. Systems in particular, became. Very unwieldy. We. Got to the point where there, was a lot a lot of manual, work a, lot, of humans, expected, to be very very basic. Sheets. Type, work, and. We felt that you, know we needed to basically improve that on somewhere we also had a lot of data and. Google is a dear a first company we make our decisions based on data. And. We were still making our decisions, based on data but we found getting, not the era which, becoming harder, and harder for our accountants, we. Were asking, them to go and. Do very very basic work, to extract that data and. Do something, with it and make those right decisions, the. Other new. That we had in Treasury in particular was that we, were. You. Know kind of like one of the last lines of defense of closing. The books and our CFO's. In. Previous, years and currently of an. Expectation of we could close the books fast that, we've got little latency, and, we can close them supremely. Accurately, really. Really fast, I'm not was getting harder and harder in, today as were all given the systems and the fragmentation, of systems that we had in the treasury ecosystems, so we felt that you, know to transform.
Finance, Meant to transform, Treasury, first it. Was our biggest kind. Of headache and, that's where we started and focused on great. Praveen. You are also talking, to a lot of clients in the market and looking, at their challenges, and what. Do you see those clients are looking, and asking to us about, their challenges, so. There are multiple challenges but if I have to classify them I classify them into three the first one what, our clients have you know facing, as a challenge is building. The business case and. Also a. Roadmap, an approach, on how to migrate to the cloud that's the first one the second one is dealing, with the massive amount of data data, is there but the, how they can in a better utilize it and have a data-driven, insights, and the, third one is around the, innovative, technologies, are popping up crazily with hyper scalars and how, they can leverage these, innovative. Technologies, to enhance, and. Make their business process more efficient, great. So so all I think I think the common theme we are hearing is about intelligent. Enterprise how, you can leverage the latest and greatest of technology, that Google, cloud platform is offering as well as si P is bringing, to the table. With. That I think would. Like to just play a video to just show you what. The future looks like with. The intelligent, enterprise and why, we are talking about the intelligent enterprise right now so can you please play the video. Humanity. Has, long dreamt, imagined. And reimagined. Itself, how. We live how, we connect, how, we can, grow. Prior. Generations. Have created, a technology driven world fueled. By imagination. Now. It's, our turn, we. Are taking the first steps towards a future, where humans, and machines will, become integrated. Our. Generations. Momentum, in bridging, the human-machine divide. Is the, achievement of a dream that crosses, time and civilizations. Until. Now machines. And humans were structured, as distinctive, functions, of knowledge, and execution. Factory. Robotics. Machine. Learning, predictive. Algorithms, blockchain. Adoption, and digital. Twins are only the beginning of, our capabilities. Designed. To bridge this human, machine divide. Applying. These new and evolving capabilities. To improve our human capacity and, productivity. Is the, challenge, of our generation. Today's. Reimagined. Platform, delivers, solutions for, improving invoice, accuracy. Sensor. Based supply, chain visibility and. Increased. Production, from smart factories, at. Deloitte. Our, search. For new applications. And better interactions. Drives, us every day. From. Our innovation, labs we, bring to life of better more, connected, world with. Our leading-edge, work enabling. Digital, twins. We. Face a critical juncture, in balancing, our time between the attributes, that make us human. Managing, complex environments. And maintaining. Our ability, to step back and dream as our. World becomes increasingly, digital, our. Biggest advantage, is. What. Makes us uniquely, human. Today. We celebrate you, the. Idea, architects. Of the future who. Understand. Our unique human ability, to imagine and reimagine. Outcomes, everyday. Thank. You, so. With that I, think. If you can talk to us a little bit tell us a little bit more about how, you are using these technologies to, make your enterprise, more. Intelligent, and and. Your plan of handling, the data growth and stuff like that so we, talked, about the. Problems that we had in Treasury. Unfortunately. I don't have a, really. Cool video to show but. What I do have is the. Power of s, AP treasury management coming, together on the, Google cloud platform, to. Solve our problems so you, know what we got. For.
Joining. Forces was. Super. Strong identity. Access management, that. You could get with Google cloud obviously. Treasury. Management, has. Got lots of dollars lots of people with the authorized. Ability, to move those dollars around the world but. Also a, concern, right, the wrong people being able to do that as well and the risk that would have on our business, so. We needed to have something that was very very secure. We. Also we, needed to be able to have an to end encryption so, no, matter where we were moving data around. We, were able to encrypt, that very very strongly and we got that with Google cloud. Also. By, putting, some. Of Sapa, technologies. Like business, warehouse together, with, Google's, bigquery, we. Wrote connectors, to. Be able to get those to talk to each other and that. Became a power, center for our dinner, needs and rapid, fast data access. Etc. And then taking. Us for Hana, unused. Trading out and usually not as a full-blown. A or P at least for our treasury management meant. We got rid of ivory. Complicated. Ecosystem of Treasury, where. We were using you know greater than 150, different products, tools and. Manual, processes. We. Were able to wrap that under one roof. And, again. This was the first, outing. Of SP. Financials, on. The. Google cloud platform we, have other for. Example we've got a tight, IT, asset, management for. Sa P we've been using concur. Also. For years but this was the first outing, for SMP financials, on, Google cloud platform and, with that came a few gaps that, we had a build, so. We've got no accelerators. Around. Some of those as well if. You could go to the next slide. You know we saw if. We take a slightly deeper look what, we built for Treasury, we. Were able to enjoy a. Bit. Of build versus by, coming. Together so. We've got internal, tools that we built in, Java that, we can. Connect. Straight to the Swift it's. Just, payment. Systems, we. Built also. Connectors. Between. S. AP s BW on our own BQ. You. Know we built those connectors, and on tools in Java as well, and. You know we were able to do that very very, seamlessly, using. The Google cloud platform. We. Got, exceptional. Integration. Near real time it. Not quite super real time but near real time, given. Using. Google Cloud storage and pub/sub, etc. We. Got great connectivity, as, well back. In the air CPUs. And some other out-of-the-box, products, like. Cloud. Connector, and s AP router etc. And. Of course one, thing that I'm very very proud of is the. Collaboration. That ASAP engineering. Organization. In Waldorf, in Germany. Built with my engineering, teams to come innovate so when, we were looking at the treasury management system. We did see some of our own business requirements. Particularly. Around hedge management that, weren't. Quite where, we needed them to be in SFP we, thought should we just build out ourselves and do a tag on or bolt on the ASAP and again given. That we started. In a fragmented, world we were trying to avoid that. We opened. The conversation, with Waldorf, engineering, and they were very, very open to take and what, they called sensible, use cases that they could scale and we collaborated, with them both, in Germany Colo Kiev and in, our offices, in sunnyville. Co-located. And they built our requirements. Into. Their core product that everybody got to enjoy in the 1809, release so I was very very proud of being, able to work with those really awesome teams in Germany to. Do that we go to next slide. One. Other thing that we. Got to leverage them from the Google cloud platform side.
Was. The infrastructural. Elements, so. For example, we were able to scale, really. Really fast and we were able to have, super, high availability by, having multi zones back in, accept, run of course our field we'd simplified, our fail overs, because. We had Clyde. Load-balancing, playing, on our favors I'm also in terms of, management. Tools to do the deployments. We. Were able to connect, our stack. Drivers. Integrated. Seamlessly with ASAP Solman so we were able to do end and monitoring. We. Were also able to. You. Know have our own, Prem services. Integrated, through. Google Cloud Storage because. Again some of our ecosystem. A lot, of our ecosystem, around core financials, et cetera et cetera are still sitting on other a or Pease and fragmented. Products so, it was important to us that we were able to connect, seamlessly, into, those on Prem products, and we could in. Terms, of. Durability. Of our data, no. More bit, rot that that. Was a thing in the past so that was something we were very very excited about as well and of course we were able to provide for the future we. Were able to have, our SMP certified, Red Hat which. Was very important to us etc. And then we talked about the security piece again, this was finance this was Treasury, this was billions, of dollars moving around the world with, lots of authorized, users, and potentially. Lots of unauthorized, user there was very very important to us that we could encrypt we could lock down. Who, was. Able to access who couldn't we could have single sign-on we could have token, exchange for. Authentication. Etc. And we got that with the with, the GCP, rollit okay. And I know, security, is a big topic for everybody. And. You. Sounds, like you feel comfortable running. As for Hana on GCP, and the security around the entire. System the reality. Is where we, we. Are a go to launch, this we are launched pattern right now we. Go live on May 1st, with. On. GCP. S. AP treasury management system, so we are very very happy we, have completed, all our user acceptance, testing, it's. An exceptionally. Strong response, we, have completed all our load of performance, testing very. Very high. Speeds. And. Accuracy, so, we are a go and we are going to cut over our entire trainers, our email management on ASAP on Google, cloud on May 1st perfect. Perfect and I'm assuming given. Your data volume, with, the announcement. Of 6 terabyte, or 12 terabyte, definitely. You are set with on, Google, cloud platform to. Run your s4 Hana absolutely. We're we're good with 3t right now, but. In the future it's. Good to know that we can extend where we need to ok, perfect so Praveen, back, to you. Similar. Intelligent. Enterprise concept, and leveraging. The power of s4, Hana as well as Google cloud platform, I.
Know We talked about clients. Are expecting, us to innovate automate. And speed, up the business. Transformation, process. Can. You talk to us and tell us a little bit more about what are some of the innovative solutions, that Deloitte is building yep, so. Actually. The the pace at which the, hyper scale is like Google the releasing the innovative solutions what. Our clients are expecting, us to understand, the technologies, and as well as understand. You. Know how these innovative, technologies, can be incorporated in, a business, processes, so that they, can be made more efficient, and in a better way and also. Help. Them with the building blocks as well as coming. Up with these solutions because, right now they're, expecting, us to you know deliver. The solution wait they can consume and rather than building by themselves, so, looking. At this pattern you know leveraging, the power of you know ASAP, ASAP cloud, platform, on Google. Cloud platform we, developed, a few innovative solutions, which in a war to showcase that maybe we could realize you, know using these technologies you can make into, the processes, better technology. Better so, that you can automate. Deliver. In a much efficient, way all. Right. So. What are the in, the technology, innovative solution which I want to talk about is. Most. Of you have done siding I think you know there must be most of you guys close to being an existing, a safety client and some of us be embarking, on this ap journey but. ASAP. Sizing, is a you. Know a journey by itself it's a very cumbersome, have. To work with the process teams to get all the inputs so that we size it in the right way even. Though you know you may need to have servers for your regular workloads, but. You, know you've cited, for, you, know your peak loads like BT or Montaigne your quarter, in or your rear end so what that means is, traditionally. In the on-premise well we buy the entire Hardware, write it and then we can have the, entire hardware, is sitting idle but it's only utilized, maybe four five days or two days you know in a month or in a in a year that's, where you know we're wasting most of the resources and the, the capital cost involved with it imagine. You, don't have to worry about s, AP sizing you just go with two app servers let. Machine, decide. Based. On the workload and everything, it spins. Off an additional app servers and whenever. It is needed, there, are additional. Capacities, available and when it is not needed, the. Whole app servers can be shut down and, decommissioned. In instances, so that way you, know you don't have to worry about the sizing at the same time don't, have to worry about you, know how much infrastructure, do I need to plan for and you, know even have to tell or I t's you know infrastructure, teams like we'll be ready within the provisioning the environment, and everything so, that's the one which I'm going to showcase right now. It's Kim. So, this is the high-level architecture so. We, you. Know leveraging, the Google, functions. And advil as Google. Machine learning technology we. Built it intelligence, and obviously the, SFA's, workload is based on obviously when there is more work it's not about the, CPUs, and the memory the. Intelligence. That we built in looks, into, the, s APIs work processes, and based, on that there. Is a machine, learning algorithm, which, kicks in and based, on it makes an intelligent, decision to spin off so, the, demo that you are seeing right now here. We have one app server and we, are in the backend and if we ran some scripts to, put lots, of load and. As. Soon as there is you know enough load that's, when the system understands that ok now I only have one app, server but, there is more and more load. Coming in in. The past I have seen this kind of a behavior so. That means you know that's where the machine learning and it comes in to rescue it, actually. Launches, an instance. This instance, was not available, before now. It spins up a you. Know actual Google instance, and deploys. Your, s AP app server and the. One of the biggest challenge was around. Adding. This tree s MLG, obviously, s MLG is a good, sa.
Piece Proprietary. Technology, but you know we were able to you, know make this possible, using. The Google functions and everything once the app server is spin off and they say P absolute. Is installed, this, app server is attached to some LG so that way s AP, can see it has a load balancing, and a load balancing group and it, can disperse our you. Know do the load balancing across the new app servers, so. Once. This is spun, off you could see this in, our SM 51 transaction, that the app server is available. And. This. Is where it, ought to be if automatically. Goes and apps it so you don't have to worry about the. In tradition, world as soon as they're in an app server is installed we. Go to the basis team saying that hey there, is racks are available can you add it to the Lord, log on load balancing crew we don't have to do that, the system will automatically, take care of it. So. Praveen, is. It basically sensing. The load itself and, at. The right, moment it, is spinning up a new observer yes and it, gracefully all flows the load to the new app server and. It. Once. The load is all coming down the other thing what it does is it doesn't go automatically, shut down what it does is it. Waits till, the last background. Job or lost long-running, work processes, is, running. And it. Will make sure that either the new requests, that are coming is not sent to this new app server and, that's. Why once. The load is gracefully, have, come down on the new app server it, shuts. Down the ASAP system and go decommissions, the. Instance that's been provided so. Maybe, for the benefit of everybody if you want to do this on prim can. We do this on prim or this is exactly, why, you should, be going into cloud to use the machine. Learning and functions available here so anybody, who has worked in infrastructure, and bases this has been a biggest, pain for all these years and we couldn't achieve this and on-premise. And this, was made possible because, of the. Innovative, technologies, that our you, know hyper scalars are providing, and especially you, know the, Google platform. Gives. Us that niche technology is that, with which and we were able to design this. Perfectly. So. Can. You talk a little bit what exactly, it is doing, so. It's. On, the app server and then, you know once it is shut down it, goes and then decommissions, the instant, that was, created. So that way there, is no consumption.
From Our infrastructure, perspective so. That's why that's how use in, a go and say the cost either, from an infrastructure, standpoint. Perfectly. And I'm sure this is not. Just for the normal days in lot of business scenarios. We. Heard yesterday a lot of retailers, have this. Black Friday situations. Where the workload goes up all of sudden and. Montane. Processing, this, could be one of those that can be quickly. Spin, up or spin down, and, you don't have to have the infrastructure. Always up and running right absolutely. Right I mean the, biggest, cases, were around obviously the. Your. Avenue, every. Friday you have a time sheet that's when you need more processing power month, in processing, and I have a and, one of my client previously, weighed, in, in, day. To day work loads all the need was for app servers but, for month end we, used to spin up 20 app servers you know which used to run for 15, days to clean the clothes there month-end and those. Extra 16. App servers were sitting idle not doing nothing for the. Stuff you know the days except. For those 15 days. Perfect. So. The other one which you, know I wanna showcase here, is around invoice. Processing solution. So, one of the things what obviously you, know most. Of you have in, the world on invoice processing as, well there are leading products as you know in terms of making sure that they, bring all things, together in the process obviously the. Invoice processing starts, off with receiving. The invoices, either it is a paper invoice or you. Know through an email or three effects or three EDI transactions. So, but, in reality most of even what we have realized it 50%. Of you know the invoices, that ap Department, disease is still paper-based. So, what that means is it, somebody, has to manually. Go enter, and then, the time it takes to enter an invoice in, approximately. 20. Minutes right. I mean that on an average but. You, know if you can automate it let's say you have an OCR. It, does the OCR then it goes, to the processing, it's everything, is good you know finally, in what is paid ultimately, the goal is how soon the, AP.
Processing, Team, can pay the, vendor and the vendor it's on time and also, the. Majority of. The time the, month end the, time it takes is to you, know validate. The GRA R so, the, solution, that what we built is you. Know it takes care of the OCR, you know reading, the paper invoices, doing. The you, know overall you know invoice processing going. Through all the validations, and also. Finally, posting, the invoice followed. By clearing. The G Aria how, this is possible is through, you know three, technologies so, what we have used so, Google. In a cloud platform and it provides us with in, a Google machine learning, vision API through, in a vision API you. Know we could achieve OCR. So, you, scan the invoice and then, it's digitized, and you. Know it retrieves the values and followed, by it post, the values into a say P you, know then it goes to the processing, then. Finally, if everything is good the, invoice, is posted so, then, the other technology, what we use was around you. Need to have these invoices, stored somewhere where, do you store it obviously we need to have the document management solution, but, all document, management solution. Comes. As as a very, expensive, one but with, the Google Cloud Storage bucket it's a fraction of a cost with, which you can store the, documents, so, all the invoices that you are receiving you, are digitizing they. Are actually, the ones the OCR is done the, actual physical copy is stored in Google Cloud Storage bucket and the. Designs, values are passed now, on to the ASAP system so. That you know the processing, can start and we. Made this solution as a you. Know a technology. Agnostic, when. I say technology agnostic. Anytime. We talk about any add-on solution, in a say P world we, think about deploying. An add-on on, ASAP system, obviously. There are a lot of versions and everything and if, you have to put this thing we have to go to a lot of change management and which, lot, of RS ap clients it's a big pain point this. Solution is deployed, on ASAP. Cloud platform, so that way we, don't have to inter, touch any, of these ap systems. Underlying, so, that way it gives us an abstraction layer all, we, do is a simple, connection to. It is an EC C or s, for harness, system where, it pulls the values processes. It pushes. It and whenever. It needs a mission learning in the intelligence, it calls. An API call to Google Cloud Storage bucket or a Google in a vision API and it, it, does do in the intelligent, processing so. The intelligence is coming from, the, Google technologies. And, the. Whole solution, since it is an SI p based solution, all our clients wants, to have an SI p look and feel so that's why the UI is kept. Unique, by, making it as honest, a plea cloud platform, and on, top of it this. Is a subscription-based solution, so, you can subscribe it there is a one-time deployment. And afterwards, as the number of invoices, you you know you process it. So. I'm going to quickly walk you through the demo. So, this is exactly where you know. We are. Processing. The invoices. So. This is the invoice that we picked. And this, could be from, any mobile device. Any. PDF. Or anything yes. So we, even, though we're talking about you know the paper. Invoices but, the another feeling, efficient of solution, is let's, say there's, a high profile invoice, that needs to be paid right away so somebody is scanning it you can place it on a desk you know take a in a picture of it from your mobile device and there is an app and, through which it's.
In A stash this process so, it's as agile, and mobile as possible. So. Once you, know it is uploaded, this, is the Google Cloud Storage bucket where. You, know you see it and it's successfully, you, know uploaded, to the Google Cloud Storage bucket now, it. This. Is an AP, processor, screen where, all the invoices the way typically it happens in you, know any, finance, department, is in a day you get in, 100 invoices, so, there will be an AP processor, who, goes the, exclusion wises puts, it in the scanner and then, his cancered, he or she scans it and all, these things would be staged the, another AP processor will go in and would, start in the processing, it here. So, once the universe is uploaded, now, the machine, learning algorithm. Picks, him now the, Google vision API is in process, where, it's actually. Retrieving. The values you. Know and, present. It in a editable. Format if. The. Invoice is known for the tensorflow, model, it, the, confidence, factor is 99 percent right. So it. Just goes and posts it but if it doesn't then, it waits for AP processor, to, go and make the corrections, let's say if the print out the, invoice that you have received the scanning is not correct it's not able to identify a value, it, waits, for the AP processor to correct it and post, it so. Now so, once, everything is all good it, goes and posts so, now from, a typical apart, from it being posted, you are under, on premise in s400, solution which is deployed on Google cloud platform and, this. Is the invoice that was posted. And. Under. Studying it right the physical, copy has to be attached to all, the noise that's been posted because it's, a compliance, an audit requirement, because at the end of the year, when the auditing happens auditors comes and say give me the invoices, that you have received so, that way you, know they are securely, stored in Google Cloud storage and these, are available from.
All Your invoice processing, transactions. And even, if you go to. Be. 0 to be 0 3, you have this you. Attach it once it's available in. Any of the transactions, where it is permissible, so. That's exactly what we are showing here I know, everything, is a theory based transactions. But our, function team some of the users still allow the old SAPT so, if you go to if, they're accessing. The transactions, through ASAP gui-based you, know they can access it through generic. Object services, as well. So. This is the same invoice, that's been in, a posted. And. You, go here this is our general public services and you. Pull on the attachment, list, and. You, have the attachment. So. What, is so you know uniqueness about this yet there are in a leading industry products out there but this, solution because, of the innovative technologies, that's. Available on the. Cloud platform enables. Us to provide, or address, the client business problem, and offer. This as a fraction, of a price as compared, to the, leading products out there in the market, yeah. And, I'm. Assuming this is also valid for any business document, not just the invoice. Yes only. Invoice, or patio it could be any it could be HR, documents or, sales orders anything, it's applicable and, we, have seen the use case into HR, organization. As well when, you, hire people you fill up a lot of forms and stuff like that you can get those as well. When. You say praveen talk about a little bit about the fraction of cost are we. Talking. About 20%. More cost. Effective or 80%, more cost effective, 35%. From. What we see in the market direction, in the cost okay, perfect, perfect. All. Right thank, you thank you for being. So. Going, back to Ray I think now. We understand, about how. Enterprises. Are leveraging, these technologies to. Accelerate, their, adoption. And. Implement. The SI p solutions, where. Do you see, your. Organization. At Google, Treasury and Finance organization, going and what you see there next nips are sure, so. Now. That we've got the foundation led and, we're implemented, in the next coming. Weeks I. Think. You, know the next obvious place to look is. Leveraging. Cloud ml. So, there's, a few things we have in mind, around. That so for example, being. Able to spot anomalies, like so. Anomalies. In our payments, either intentionally, or unintentionally, etc. I'm getting, ahead a lot and training some models to help us get. Alerted to those patterns quickly, and getting ahead of them. Another, one is obviously. Using. Advanced. Forecasting. Parts, of ml -. You. Know predict, our cash position better so how, are we using our cash. If. It's sitting in the bank can we be using it more smart can, we predict, how much is going to be sitting in the bank at any given time over a longer period of time to, get.
Some Sort of insights, about what. We should have in their bank and plan on what we can do with that so we, could be a lot more efficient, with our cash if. We're able to do that and then obviously your, trading pattern seems like an obvious more they know a lot of people have tried to crack this before but. We would like to start somewhat. Simple, on what I think is simple, by. Understanding. Foreign. Exchange trading patterns, so at. Least internally, within our own company, so using. Our in-house banks around the world can. We use, cloud, ml to predict, what. Some of the forex needs will be in various different locations and, ensure that those in-house, banks are sufficiently, supplied with our currency, for, use when we need it and, I think we can we can do those rather easily one. Other thing that we're trying to really, think about is, next, generation business intelligence, so, this. Is the whole concept of delivering. The right insect at the right time to the right user within the right workflow, so I don't want people to. Have to exit. S ap when, they're in the process of doing an accrual for. Electricity. For example I'm, trying to understand what should I have crew they. Leave the workflow they go find a dashboard they change the filters they do whatever and five, minutes later they've found out what the average. Expenditure. On electricity, is for the past six months that's. A great number you read it opposed to note you go back all the way to your. SMA, screen again by then our, authentication. Is so tight and made of logic vote so you go back any and your log back in you start writing the accrual again what I'd like to understand, is is there places where we could surface that insect, within, the, workflow. Seamlessly. So we don't have to create these gaps. And, exits. And entries and. I think there's there's a lot of. Cloud airmail we could leverage to, do, that so that's how, we're thinking about some of our next steps okay. And. I just want to let. You all know that s AP and Google, antelope we are doing, a lot of things together in. The market for our clients, not. Only building the innovative solutions, but but also looking. At how clients. Can solve the big data problem. We. Also understand, that every. Problem is not about just bringing a bigger Hana boxes, right so looking at the business at. The core and trying to solve the problem we, do have certain clients, who are talking about 32. And 35, terabyte, of data today, and moving. Them into cloud and continue to grow with 20% year-over-year, is, not the solution right so. We are looking at all of those things. Sa. PE and Google in the light we are also working to help. Our clients to to, innovate, new things, sometimes. It may be just the beginning of starting to migrate, either. Running, their POS a ps4, Hana we. Are also exploring options which, may in future, leverage kubernetes, and containers, to, to, sort of host. And run an ass ap solution, and how. That can be portable across. We're. Also excited about, pantos, and, live. Migration how. We can live into a multi, cloud environment, which is in my, mind was one of the great step that Google has taken to. Bring our. Client, environment and think about our, client, situation rather, and you know our own situation, so, those are pretty. Exciting. Stuff that we are all doing together.
2019-04-15 18:13