Subscribe
BTOES Insights Official
By
February 16, 2021

Enterprise Architecture Live - SPEAKER SPOTLIGHT : TOP AI trends and approach for enterprise adoption

Courtesy of SAP's Rahul Lodhe, below is a transcript of his speaking session on 'TOP AI trends and approach for enterprise adoption' to Build a Thriving Enterprise that took place at Enterprise Architecture Live Virtual Conference.

sap_vector_logo_small (1)

pillar%20page%20line%201

Session Information:

TOP AI trends and approach for enterprise adoption

As we know the success rate for AI projects are very low, many use cases reach till successful PoC phase by creating ML model and desirable outcomes based on sample data; BUT moving them to production with real data starts coming from various sources ex, S3, ERP, Social media etc. and then the challenges start.

With changes in the model during training, accuracy in Inference goes down, versioning of dataset, model explain ability becomes important. This session discuss systematic approaches to take for scaling AI use cases.

Session Transcript:

I am thrilled to have our next speaker, from India, to the world. We are welcoming Rahul Lodhi, who is the Senior Director for ASAP Artificial Intelligence. Hello there, I will great to have you join us today. I'll do a quick Bio on on the Row and then let you let him leave the proceedings for you.

He has over 20 years of experiencing and designing, developing and testing enterprise software products.

His areas of expertise include analytics, business intelligence, machine learning, and AI.

In that Sap, he's responsible for developing the Machine learning and Artificial Intelligence Platform for realizing and scaly intelligent enterprise grade industry use cases is also built the City Scale Epidemic simulator for covert 19, used for prediction of covert 19 cases for city administration in association with research institutes. He's working with the NASCAR covered in 19 Industry Task Force to build India's covert ... data platform to collaborate with various state governments of India.

Real honor and pleasure to have you as a row and the looking forward to your presentation.

Excellent. Thank you very, very much, ... for that.

Excellent intro, and I'm honored to be part of this podium.

Always, I was there last year as well, embedded, very interactive, audience, and also, I would love to be this time. So, this time, we're going to talk about two important topics.

Screenshot (37)-2First topic, is that, how AI is transforming in 20 20, and how do we look at, you know, the chains of AI in 20 21, so, typically looking at, got, easily be the evolving topic, and it's been a hot topic all over the day date. P C, the binary came in, and then, there has been some kind of change the way we choose to use at .... So, we be, Let's jump into the topic, and see that, are now VP takeaway.

So, what we've seen in during the pandemic time or prepared and dying, right? We always used to talk about many of the AI, it's come on board. We had, you know, we need to need some three years of data. We get five years of data to detect, you know, next one week of prediction.

What does happen during the pandemic day was we always we're looking forward for we didn't have so much data if you need to look for that, how does my supply chain? Because it was making, you know, lockdowns different states we do not know how the widest population happening out there. There has been also demand supply variation. I would say, the topics which had not been. So what's ... or some other topic where you don't learn little, We take it back.

So we didn't have so much data to perform the prediction at the fastest pace, and that's where the division started coming, that we are trying to look at the small batches of data based on to make the immediate prediction.

And there are many new algorithms that develop during that period of time to understand, you know, how I will just take an example of how the vaccines curtailed upright.

Or how we did this prediction of how does the typical, no progress of the virus going to happen is just based on a couple of months of data, which we predicted to the analytics to be, to be made.

And that's something you could type, process, that we might not need of your site, and get data, can be use such a way that it can smaller set of data. I can still get a prediction to take my job. So that's one of the I think a lot of work happen on that site.

Btog CTAThe one important that impacted all of us being, I think, Technology and Cloud.

When adopted very heavily, we see the interaction with the end user will be changing, and that the interaction with the end user is now more coming. I mean, earlier, we had more of the computer that that we're using to click through and do some stuff that we had a phone calls with, their knowledge, we will take an Alexa, Google Play, and and many other devices, which we just need to combine to create activities for that.

We're looking forward to somebody, because we know, also been adopting this technology. And believe me, I mean, going forward, everyone wants more comfort, and you don't need to be on their voice deep.

And the Chat pod, or the conversational AI, has become a very, very prominent, trying to pay, a lot of innovation happening, and they've been developing very, very ... updates there. These are not the ..., somebody is going to change our interaction with the system for interaction with the computer, or maybe we lower our user interactions, all user interaction, going forward to the conversation? We and more. What is the manner?

Important part is robotic process automation or options, so if you see, there have been always, We had different trends and data streaming with sufficient.

OK, there, we are trying to, Automated many of the maintenance task, or Nepal wealth must become.

Someone just said, you know, I want to create a purchase order in putting I want to procure something and then just the milk of that, I want to buy a laptop configuration. Now, typically, the ITK would transform the invading the purchase order. What happening is that, the company itself is really that the AI and creating the purchase order, the help of that particular employee. So, like that, there are many, many disruptions happening where the mundane task where the text can be read by a computer are being used some machine learning activities and automate that task and perform the business. What would that be?

The resistance of the automation has been becoming very, very optimal, and a lot of innovation happening these days.

Another important trend that I see, also, because the fact that we are looking at the way data that we get to circulate the data, the data, it says, ..., and this is the trend that we see.

It's been very actively taken up by the IOC industry, stakeholders, as well as data bit of input, without data, would a prediction, or on making the entities, a lot of tools, technologies, even in the algorithm be building, you know, in order to make the data more accessible, and the quality.

Next one, also the two phase with thinking or already networking, but yeah, you can be using a no right or wrong and, you know, once we take the biases of the AI, so many Goodman has been working on this. That how, how does the you know a responsible AI ethics?

And we get adopted because now, it looks to the emerging pain, but down the line, This is something which which need to be fairly going to really, when I did the glacier goes off, goes all the technological. We cannot leave the machines to go on ourselves to see those machines can only be biased. Or can we train based on the data that we feed into it?

So, if we have a young kids, the, it's a, point 0% of error in the data, or biases in the data, it would give you more designs on such and such. A potent ...

on, on AI, that you have, sometimes it can be busy with already.

CT was happening the NDP's dead bodies, with the port, is right, What is wrong? Because it used for businesses, because you wouldn't be the product, or, you know, that the data being a product. So, this is a topic, which, which has been divided lot.

There's still not a lot to happen in this site, but anyway, this is going to be a lot of the training that we see in the parking lot for integrating and also, continue to make it to the some logical point. It would be great.

But it is a topic that I think everyone, should we care about, going forward? and the AI come to picture.

Let's look at now as we move toward the enterprise, the website and how the challenges of adopting A better option, and this is something where we hope that we should be.

We're looking at the other trend that we see so far, how I can bring that back to the enterprises and all that. Enterprises to get working on that.

Yep, so importantly, when we talk about enterprises, well, how they can be getting efficiency, innovation is very important for them, but, at the same time, there has been, no change is constant, is that you definitely agree with me, that every enterprise has to focus on continuous innovation, that we don't opt into them.

Copy of Email Graphic Virtual Conferences (3)There has been dealing with the ecosystem to the $5. There is the suppliers there.

The way the DVD now bringing the renewable social media and so, they fix it down for all the channels, very, very important to make it there. So, not a lot of this topic with constantly changing at the technology as the teacher, right? And you can't get away with the complexity, the complexity that you see that loaded beyond visiting scholar award and we also have this multiple hyperscale downstairs potent. I did provide multiple tools we use. Google can't get. So, he's been making more and more evident that the best and people want to run their businesses or depletion. And that makes your changing process much more complex. They can bring to the other expenses they do with the UI as well, the same time your disability has to be nice, because today they are being driven different things in different systems.

But how they can get together and do the base of automation or the base for the innovation, make your organization. But so we. Are in this world of constant change is the speed.

It's been increasing. I would say every time that that new technology or new description comes. But this something that you need to understand before we understand how the AI space yet, Same time. The DEA now it's been rebuilt in the businesses.

So as we discussed the poor impact that we see going to happen also the business processes, or enterprises, in this, we can make it disappear, We see how the prediction of the poor deep learning and machine learning capabilities come up.

How does the intelligent automation is going to play?

So, that's the IP, is that, we discuss about they tried to deal with the going to be more envoys, interactivity with the system and people more and more than that.

But, important is that, how does this, this business going to factor, that, we see that human interaction task is going to be automated by 60, 60, by 60% by two pi.

Also, there are about 93% on the two, people are going to be p.l.d., personalized experiences, and also this whole renewable energy, which is predicted, that is three point nine trillion, but it's going to be more than that.

Also, the, the potential of digital technologies that we talk about, transform businesses, item. Where do we see the opportunity come when you talk about the enterprises?

We interpret this, typically, you Don't, You can be, it is at least 50% of the large, global companies will be using AI. And then designate one thing that they may need to decide how they can use analytics as well as, you know, the IOT as a part of their supply chain, which already happening.

And it is not so far away, but it's going to happen as well.

Got out and McKenzie suddenly they took care of me, is almost 60% of your does involve in the source to pay.

Processes have the potential to be fully organized, the automatic and that's where you see the lot of the section 54 departments RPS would come up there.

Yeah.

Smarter as well, fintech industries and the finance is yeah, it about 50% of overall time of books will devote to it.

But yeah, the publicity of the data and the potential of leading the automation is very, very high and it's still there.

Think stupid, you know, that pain, speech, which comes from the data source, that, that's where the technology with. The limited.

When you talk about the AI, which side of it looks good to be really a lot of potential, and a lot of, I would say, there's a lot of important aspect sleep in your businesses, forward.

Implicit in businesses could transform the piece.

What is more important also, is that a duplicate the sticky and how does the EIB, when X equal to interface With typing? To be very, very fact, is that AI is been, yeah, many of the year to date is being phased out.

You don't making AI to the production, is, is very, very less, not one seat out of the issues being dead.

At the same time, many organizations also struggled with the complexity, because this complexity will be different distributed environments. How does the struggle to get into the right kind of idea that you can focus on, you have your on prem system, you have your cloud system, your data streams, as such, need to bring them together? In order to do that? I have a project.

And another important point is that there is a shortage of data scientists.

As for the, the survey said that there's going to be a shortage of, at least, want them to be held in data scientist, which, which we would might be getting into it. And I think it's been a big change, but we see a lot of it so we needed to build the project. It should be like in an orchestra?

You need to have all the things together in order to make the right.

They exchange, not that you're just going to disrupt the business processes by not having your base at a property, and in order to make the description, you need to first think about probably on the machines your business processes.

So first and foremost, in order to hit the maximized value of your parent as a unit to kind of address the challenges to be measured in your VP's work on the current the EI into holistic way.

Before you think of, how can you get into that mix?

And this is something that, that you should start to pick towards to think that, how could you get magazines of your data, right? So your processes.

And curate the diverse data. How do you transform the data into the usable assets? Operationalize your machine learning. That's something which is which is important. So your data, that's something process. And I think that does it, and somebody readable format.

Next, important part, that you have to simplify and give your data. So Analyze, now, correlate.

That was the multi-phase data across different landscapes. And some of it can be in the way that it can be accessible to your developers and I think for me.

Otherwise, if your data is built into the different systems to pick them up from there and use it for a part of this third, you have to look forward for your implementation of AI competently.

Screenshot (4)That will get you build that environment of where you are creating and deploying the most important.

At the same time, you're making the desert that get out of your AI, has to be kind of excellent negligent, or are, you know, you feel confident that OK, it's not a bias designed or get the right amount of accuracy of the design. So, you need to build that environment, and, I would say, platform to make that thing happen.

And, importantly, the automation and the scale of the important part. Because, what can be today, is gonna be looking for a smaller problem, but looking at larger scale, How that can possibly happen.

So, many projects, you see, it has been, you know, doing good to use the stage, but, the scaling part is really, really important.

Fourth, an important pillar, which is sibelius compliance part.

So, we have been seeing the data protection privacy and convenience, the challenges, which which typically comes here is very important that determination policies are you would have also the governments around the data out.

Those has been also been taken if you think about cognitive entries compared to the fees can achieve your enterprise interpreted stateless for the project.

Other point that we should also be considered is that different Geek, but so much involved at the exit, there will be not for the IT point of view, which has been at the end user.

But according to the business, you need to have, that is the CIO or the non relational data scientist, and the operations are always doing the machining operations.

They would like to have a lot of new open source technology to be used, and their need is to, they need to know. To give too fast, is the technology.

The same time, the guy who make it into the production, he look for more compliance by, you know, what, you're making open source security, Look for life, It has to follow some security or deadline, and that's why they need and disagreement about what data science is pretty normal.

And same thing, the cost and managing document, because these are the three points from our big challenge process.

Just getting into the data, into the fact that we can also understand, what is the enterprise.

Before we begin to sort of shift, by the way, that hope we can make it more synthesis, the interface AI is uses a different flavor.

We think that something that we want to go, you don't make it very complex. But just that, it shouldn't be always hard to, from what is this? But it also should become one result business folks, as possible.

You can do the POC. So typically, what happens is that, when the ..., what we say is, people are reading POC, the lakes and data UV.

prediction out of it should be your management. And management says, oh, it looks cool, Let's let's implement data production.

You try to take this, the production. You can see the challenges that your data is not connected, properly, or marketing, right? Amount of data, because this is, not kidding, showed that the utilities be not managed. You see the challenges. When we look at the enterprise readiness of AI application, should always think about how this, this particular experimentation going to enter the business. You should look for business perspective. I did, the output of your AI, has to be go into the way that you are making to the chatbot. Consuming it in the form of the conversation earlier, on your peer, getting, useful on some sort of really getting feedback on, maybe it would be the analytics that something is going to be useful.

That's where your value of your AI would be artificial intelligence, like foundation useless.

And you should start from the data required to reach to the end stage, and I think you'd be, then, when you create the expectation and your models for machine learning, or the accuracy of the model. All those parts has to be managed in the form, that development process.

Those are not the E part impact is on your data coming and how good are you unable to produce the output.

Which also is important that the Information Management System, that we built it from the data perspective, is very, very important that you discover your data, half of the deployment of the data happens, at the same time, how you walk us through the data, because your data is coming from multiple data sources site.

It's not easy that, which will be lucky if, let's say, you did a great solution by just getting the data from one place. So, that's the easy task to do. At the moment, you have a data from your transaction data, from your on prem system, you have the value of theta coming from a different system and there are different composition of the online streaming data that you'll get.

So, maybe there are multiple Datas are impacting your overall system. It's important to also orchestrate them together and also the New York City.

We need to get the right output speed for the EIA to consume. So, you get into this aspect before the start. You know or you're working and importantly, also Goldman Sachs.

So, you really need to think about that, how does your EIN, we use the form of the assembly line.

So, your data is coming connection, but structured data, and unstructured data, or the hot stream data, which could be, then get into the orchestration process. So, you will have attestation as exit of the patients, machine learning, and deep learning team to make it. So you basically get your data on secure data. You build your model, and you bring your deployment.

Where do you have to deploy the model?

The deployment would then look into the, how does your model performance cost.
22:40
Your model is as good as the B, It could project the data as and put as the event. And you should also strategy about how do you replace the model?

Screenshot (37)-2And you don't get a point of time on it as well.

The scale part, as you've already done, being forward for, let's say, I know, build the model for when division one can be modernized to scale to the, you know, multiple countries are on multiple, the weekends or multiple distribution. So, that process, as we didn't talk about enterprise, is, it tends to be, the scale is important.

And you should move forward for somewhere where the ... maintenance can be really easy. Scalability plus, K does not explicitly.

one of the things that you're doing, but one experimentation and Berlin, when you get more data, more variation of landscapes.

So, this part is very, very important, and you have meeting the machine learning model.

And, last, but the very important part is the whole year.

So, the time to consumption strategy is not there the use of the opening of mini grids, or getting ready for the output of the other one.

Overall experimentation is important.

That consumption should be either intuitive, but that appears on the analytics on, strengthening back to the business processes with something which to make the automation possible efficiency by using the UI. So, any business case that you have for AI into this assembly line concept.

Then, they wanted to start your project on. Otherwise, there will be chances of either you that you disconnect this piece, and you will always be end up making something, the basics, if not working on a time.

Your pipeline is not getting data, or, an extent, is, not getting your application, then you will have some challenges. So I say, the architecture point of the ... point of view is a very important phase, that you should focus on.

And this is the way I think we're going to consolidate these different personas. We're kind of working towards it or the the ... that always is showing. They would have of Lincoln's Inn Katie of Inquiry cycle.

And you also know that how we can update it didn't it?

Legion that if you how is part of your site, they get onto the Costco Not only manage this, for instance, with the big example. Pick it up, Yes.

So this is a simple example that the red example of this I can. take that concept.

That's something we're talking about, It's called plug in from the production manufacturing plant.

Then you can think of somebody's been getting this done for car manufacturer.

Got a good manufacture that I would say is important part is that one, this particular component can develop and to show that, that I call it in this particular process.

Teens eat by either using BA.

And how does that happen here?

This particular company, they had the, I add images being input, So as soon as your company is ready, you should be taken.

Typically, with the machine learning, this section, you can see the dimension of the base.

You already have data, and that data, you try to match your Exit OK tonight.

Then, you'll use the IOT sensors, So, IOT sensors, and kind of your information on it, and the sense of safety with that.

It's been lately.

Third one is that can fit a master data of this particular component. So you can check that ... system and get the study now find that they have to make this application when we say that.

Yup. It's the quantities, right?

And also the how can I predict this can be done data and also in a Smart Quality app.

They meeting the police forces, so, you know, they just don't know.

So, if you're just looking at the architecture of high level, concept point of view that, that is a manufacturing, you'll now have WordPress.

And then, the IOT, typically, central, Kafka, and Kafka would already have incision you, which stored all the mainstream.

And then you have images taken by Deputy David cicilline influence or data.

But in order to bring them together and again, the three bits using hana and the LP concept here, you can connect them together at the same time. You don't need to get things for production, very. Those things has to be joined.

But today, if you're not thinking through and define each polygon, simple problem, you might have really put your solution. But it's not connected to your golf cart.

Always, not going to keep your back end system, your mental challenges. How do you show this?

You're yet the app.

What has been I think the re-imagining layer is the theme to the problem from the beginning.

Copy of Email Graphic Virtual Conferences (3)

So first thing that you should do is that we have to make these data to be also the data lake.

So to pay the data pipeline where you already have your Python model created, for example.

And which is really combating the I had images, versus the enemies that right, or wrong.

And then only these data, the data, and also use the algorithm for prediction. And that will penetrate.

And this is the way you, being the, the end to end process, though, in greenery band input equal, to produce the impact of business process, and Sydney, the predictive ability to deliver the right on the regular basis.

So, this is one of the example, that, that will bring, but invention in somebody to do the kind of environment.

When we talk about bringing the beginning to a landscape, or on your enterprise getting, navigating this, first thing, is, that your data is something for your data, can be done by the stream data, on the structured data, unstructured data, of your spatial data, you need to have a right engine on the machine learning like this.

On the one hand, there are many other open source, is to embrace the open source thing, because on this technologies has been buzzing with open source so far. Is the key to it. You don't open source. You don't.

You find out that you can't re-invent everything by your own, and everything has to sit on your cloud environment to be done on the dog on the stack. scheme. So and then when they're there is something which important. Required.

Has to be scaled HIV on.

the therapy that in France, Jefferson, become a bottleneck. And this is the way to move forward with the magic here.

Just to put this in perspective, how should you start the legitimate typically, should start your journey with the young, making all the data together. Make your information, manage that environment, management part ready.

Make the delivery of a comprehensive data and digitization for the next part is also new.

Visualize the data and, and look at that from the BI solution to the limits of what we'd like to move forward toward modernization. And be able to at least get these data sources.

Right.

And then, you should look for each mutation.

Machine Learning is use case that we bring it up and followed by.

You should look for the deployment of a machine learning model, deep learning into the production usage. And the important part.

An augmentation of the US a business process is not going to enjoy a complete clarity.

This is the way you basically see both, how we want to bring the enterprise readiness to the BI solutions.

And so you would able to never reach to make it to this existence of the crux of it.

Is that one thing that you see, to make that successful company, the assembly line approach, audience, that helped you?

With that, yep, that's it.

You so much for these things.

I'm going to Josh, if you have some questions, are more than happy to address them.

Fantastic girl, very much, appreciate your presentation, Your cover, your cover a lot of ground to cover. So much related to enterprise AI, right off the bat.

Let me ask you a very practical question that has emerged, which is, I mean, you show the level of complexity, that is, that, that, that, that's almost required currently, should have good enterprise AI.

And, I have an understanding of all those components. And, then, you also talked about the need, just the simplicity of some aspects of it that they need to simplify some aspects of it, so.

So, my question is, you know, since we last talked, it's being a few months, since we last talked, and, you know, AI, Enterprise AI continues to evolve, you know, the current state, what are you? What's getting you excited about AI at an enterprise level?

Things that maybe, you know, 1, 2 years ago, were not there, and now you're thinking it's not hype anymore, it's ready for prime time, what are some of the things that you're excited about when it comes to enterprise AI?

Yep, I think, good, good question. And that's something I would literally, think about it, as well, as part of my job.

Do you excitement at that level?

I think a lot of things has been transforming, as well as AI as a field.

And we see there are the trends, which have been changing quite a lot, especially when you prepare the enterprise grade applications.

The important part is that the business processes, which are the key to make over our business, or enterprises more smarter, I say that and there would be, the technology are either be playing very, very happy to receive any algorithms, which are coming at the bus, stop it at the same time, the bees for EI is the data. Those little bit, one of the cheekbone on the innovation that we spoke about, this coming Monday.

I use this data access, and that's something that a lot of innovation happening at that stage, where we see the quality data is coming out, which is making us to build a prediction easier.

And that piece of data, ..., ... pipelines, is we're making many possibilities achievable for forest, and also, we were just thinking about, Oh, I can make it. You know, It's a cycle, time, something, automation easier, but now we're looking at the new business avenues with such kind of opportunities which we get something which they were being taught before. Especially now, be joint Is EI completely ....

The way we see the home automation, is, it's coming into picture, as well, Business process can be just tinkering with the from directly from the chat pod. And you see the supply chain, we trigger. something has been delivered at the base of all this, into a cycle of the automation, you followed by EI, that something was, I would say, not been having the imaginary. But, we see a lot of progress happening. to make this all complex systems together, in order to build up smart algorithms and making it to the next level.

I go to the next big thing, which, which would excite me as well, to look at how this whole autonomous vehicle concept going to be G one of N.

Now, this will be the only reality, because becoming reality there and slowly. I believe they're the ones who are going to guide the next generation of the AI, which maybe we get back with them.

even part, or, or just to be there.

Start really, kind of ...

that everyone will have their own, but, you know, the robot who will assist them, and I'm sure that down the line, find something like that.

Very true, very true. And the row, a lot of a lot of our audience today, globally, are, you know, it's made up of enterprise architects, and there's a different level of maturity, Some of them have working, very large corporations.

They have develop, enterprise architecture already. Others are kind of building as they go along. And especially for that group, that's, that's, that's building enterprise architecture currently. You know, how do you account for AI as, as you start building blocks for enterprise architecture? What would be a good approach for them to take at, so that, you know, the transition to AI systems is more seamless?

Excellent. Yep. I would suggest to, you don't take always a step by step approach. Because as we see the presentation, if you tried to do something drastically at equity with the highest level of my personal experiences. That work.

3 to four years, building the field and working with many customers globally now very, very closely to make be successful.

Whenever there is something to hurry, and not your backend is the assumption, is this really busy, high chances of a particular way.

Screenshot (4)So we had to aid is not, an option is not like, building any business application, is not ideal. Yeah, you have a requirement that supplement according to your obligation ready In couple of days. I think it's always the end to end process.

We're here to start thinking of, first, I would say, I think on your database database is the is the key for it where you would take care of your code nuns your data labeling. The ... kind of a concept that would be later on be lead to big query.

Happy is first and foremost is that data pane is important. There are a lot of documents of Augmented. I'm more than happy to put any of them with my knowledge. But again, that's where you should get started, the more strongly will be tapping your CRN adoption.

Second part, is that, then, build the infrastructure where it will be, it will be more stronger when you get able to run day, Make the data available for data scientists. They could be the algorithms around it, to make your courses and the consumption of AI.

We have the consumption, make the EPA's ready environment, or in front application where they could consume the EPA's, which coming from.

So, you wouldn't mind whether independency out of your system, because it is always processing the Liberty Henry ... will run the AI within the same system, or say, men, who, I mean, definitely. It will eat up your complete space, and then we don't, and obligation. So it separately, but consume it, maybe the same location, but I would say, on a particular signal, should there, but consume that your business application of Europe, the Application Mccready, where they can do, just take the, you know, imprints of the data because of the model that you're getting into, and then delivered to your application.

So, these three layers that are very important, every known as vertical composite food, and those, to the, to the hospital and you want them happy, not supported decisions.

That's excellent.

And we have a couple of minutes, I'll sign a goal, a little bit deeper on the, on the, on something you just talked about, which is this, even on your use cases. You have some examples where, where you show this, this integration of business process, and then with machine learning, and so, what are some best practices best approaches to do that integration of business, process with machine learning?

Yup, so, importantly, now, we have to look at the, the machine learning output, right?

So, machine learning output is always, would be that, in terms of the either you would build your bureau, robotic automation, or duplicated output of this is going to be your application embedding the link or emitting the dashboard into your application.

And this integration point among these two, as do we go with the open APIs. So, the integration among these two will be destroyed, by the same time, the context as to be possible.

So, the first part of integration is security and didn't management, because that's where we talk about the, you know, whether you are into the privacy is important part. And, then, the more you're going to the cloud environment is always a tendency, is become the becomes, so that your tenancy that you maintain it or do you maintain your application? They were these application layer, it has to be followed to the URL is basically your data fit as. But Social Security and Defense is something that the competition portable taken. Second part is to make your applications as well. Total consumption should be zero from the point of view. And the scaling part of the application. From integration point of view, when the most applications come, there is always the domain entity. More than that we talk about.

So the entities that you use in your application has lots of phone to your data. So, these entities has to be also master whenever using the 3 or 4, I'd say, the best practice we follow when you're thinking on the integration security, when JP Morgan.

The Big Bang theory, this there, and from the application point of view.

So, but it's indeed a good question.

And, I mean, since we're less time, but I could talk about what's on these topics designate, yes. I'm so appreciative of talking to you about this topic, because it's always a masterclass on the practical applications of Enterprise AI and you do it masterfully. So, thank you so much today for sharing your expertise with the world, and in the end, we're very fortunate to have you provided those insights to us. And they're grateful for that, thank you very much.

Thank you, Thank you, it is my honor to be part of, uh, this confidence, something again in the beginning, sharing my experiences, and it's really nice to see you again.

Thank you very much, ladies and Gentlemen, ..., director of sap Artificial Intelligence, a real, knowledgeable leader, and practitioner of artificial intelligence, not the hype, the real practice of artificial intelligence in enterprises around the world today. So a real privilege to have him with us, and that has set us up as based on his answer, is as you, as you heard directly from him. To the topic that will be the focus of our conference next week, next week. We're gonna be kicking off Robotics, Process Automation and Intelligent Automation Conference, and that they know that was a little bit of a preview of some of the topics that we're going to go deeper on next week. Now we're going to wrap up this session and we're going to start fresh session at the top of the hour and the last session of the day. And it's about intertwining the nature of digital transformation, enterprise architecture, and secure of business operations.

And this is going to come from a leader and a very experienced and successful leader in the biotech industry. I will have English as Money who is the Lead Enterprise Architect, and Senior Data Scientist at Biogen. Share his expertise with us on how he is. He's doing that with Enterprise Architecture.

So, very glad to have you all with us. We have one more session at the top of the hour. Very much looking forward to seeing you there again. Thank you.

pillar%20page%20line%201

About the Author

more (62)-4Rahul Lodhe,
Senior Director SAP Artificial Intelligence,
SAP.

Engineering Leader with more than 18 years in world class company having extensive experience in designing, developing and testing enterprise software products. Expertise in domains like Analytics & Business Intelligence, IoT and Machine Learning.

Responsible & accountable for driving global missions with Product Development, Processes and People. Managed end-to-end life cycles of onPrem and cloud enterprise software product components – from whiteboard designs to maintenance releases.

Experienced in building ground-up world-class teams, as well as in driving transformations of inherited teams. Building the team with Innovation as DNA and transformed employee engagement and trust score. Extensive experience in working with geographically distributed teams. Managing a development team of Managers, Architects, Product owners and Program managers.

Excellent balance of Technical skills, Managerial skills and Entrepreneurial skills.
Thrives in a 'start-up like' work environment.  

pillar%20page%20line%201


The Business Transformation & Operational Excellence Industry Awards

The Largest Leadership-Level Business Transformation & Operational Excellence Event

opex_assembly

business_assembly

Proqis Digital Virtual Conference Series

View our schedule of industry leading free to attend virtual conferences. Each a premier gathering of industry thought leaders and experts sharing key solutions to current challenges.

Download the most comprehensive OpEx Resport in the Industry

The Business Transformation & Operational Excellence Industry Awards Video Presentation

Proqis Events Schedule

Proqis Digital

Welcome to BTOES Insights, the content portal for Business Transformation & Operational Excellence opinions, reports & news.

Submit an Article

BTOES UNIVERSAL GRAPHIC - NO DATE.webp?width=1200&name=BTOES UNIVERSAL GRAPHIC - NO DATE
ACCESS 50 VIDEO PRESENTATIONS
Access all 75 Award Finalist Entires
RESEARCH REPORT 2021/2022
BTOES AWARD - NO DATE
BTOES UNIVERSAL GRAPHIC - NO DATE
Subscribe to Business Transformation & Operational Excellence Insights Now
btoes19.png
png
ATTENDEE - Proqis Digital Event Graphics-2
ATTENDEE - Proqis Digital Event Graphics (2)-1
ATTENDEE - Proqis Digital Event Graphics (1)-1
png

Featured Content

  • Best Achievement of Operational Excellence in Technology & Communications: IBM
  • Best Achievement of Operational Excellence in Oil & Gas, Power & Utilities: Black & Veatch
  • Best Achievement in Cultural Transformation to deliver a high performing Operational Excellence culture: NextEra Energy
   
Operational Excellence Frameworks and Learning Resources, Customer Experience, Digital Transformation and more introductions
  • Intelligent BPM Systems: Impact & Opportunity
  • Surviving_the_IT_Talent_deficit.png
  • Six Sigma's Best Kept Secret: Motorola & The Malcolm Baldrige Awards
  • The Value-Switch for Digitalization Initiatives: Business Process Management
  • Process of Process Management: Strategy Execution in a Digital World

Popular Tags

Speaker Presentation Operational Excellence Business Transformation Business Improvement Insights Article Continuous Improvement Process Management Business Excellence process excellence Process Optimization Process Improvement Award Finalist Case Study Digital Transformation Leadership Change Management Lean Enterprise Excellence Premium Organizational Excellence Lean Enterprise Lean Six Sigma Execution Excellence Capability Excellence Enterprise Architecture New Technologies Changing & Improving Company Culture Agile end-to-end Business Transformation Execution & Sustaining OpEx Projects Culture Transformation Leadership Understanding & Buy-In Lack of/Need for Resources Adapting to Business Trends Changing Customer Demands Failure to Innovate Integrating CI Methodologies Lack of/Need for Skilled Workers Lack of/Need for Support from Employees Maintaining key Priorities Relationships Between Departments BTOES18 RPA & Intelligent Automation Live Process Mining BTOES From Home Cultural Transformation Financial Services Customer Experience Excellence Process Automation Technology Healthcare iBPM Healthcare and Medical Devices Webinar Culture Customer Experience Innovation BTOES Video Presentations Exclusive BTOES HEALTH Strategy Execution Business Challenges Digital Process Automation Report Industry Digital Workplace Transformation Manufacturing Supply Chain Planning Robotic Process Automation (RPA) BPM Automation IT Infrastructure & Cloud Strategies Artificial Intelligence Business Process Management innovation execution AI Lean Manufacturing Oil & Gas Robotic Process Automation IT value creation Agility Business Speaker Article Systems Engineering RPAs Insurance Process Design Digital Speaker's Interview data management Intelligent Automation digital operations Six Sigma Awards thought leaders BTOES Presentation Slides Transformation Cloud Machine Learning Data Analytics Digital Transformation Workplace Banking and Capital Markets Data Finance Professional Services Education IT Infrastructure IT Infrastructure & Cloud Strategies Live Blockchain Interview Solving Cash Flow with AI BTOES White Paper investment banking Analytics Insight BTOES19 Consumer Products & Retail Enterprise Agile Planning Government Operational Excellence Model Project Management Algorithm Automotive and Transportation Banking Business Environment Digital Bank Enterprise architecture as an enabler Hybrid Work Model Primary Measure of succes Relationship Management Sales business expansion revenue growth Adobe Sign Agile Transformation CoE Delivery solution E-Signatures Electricity Global Technology HealthcareTechnologies Innovation in Healthcare Reduce your RPA TCO Transportation Accounts Receivable (AR) Big Data Technology CORE Cloud Technology Cognitive learning Days Sales Outstanding (DSO) Logistics Services Operational Excellence Example Risk Management business process automation transformation journey Covid-19 Data Entry Digital Experience Digital Network Digital Network Assistant (DNA) Digitization Drinks Effective Change Leaders HR Internet Media NPS Net Promoter Score Program Management Portal (PgMP) Sustainability TechXLive The Document is Dead The New Era of Automation Automated Money Movement Banking & Financial Services Biopharmaceutical Blue Room Effect Building Your Future Workforce in Insurance Business Process Governance Capital Market Creative Passion Digital Transformation Workplace Live Digital Workforce Digitalization ERP Transformation Finance Global Operations (FGO) Financial Services Software Frameworks Hoshin Planning Human Capital Lean Culture Natural Gas Infrastructure Natural Language Processing Organizational Change Pharmaceutical Pharmaceuticals & Life Sciences Project manager Supply Chain Management Sustainable Growth The Fully Automated Contact Center Transformation Initiatives Workplace Analytics eForms eSignatures 3D Thinking BEAM BFARM BTOES17 Big Data Processing Business Analytics Business Growth Centralized Performance Monitoring System Communication Creativity Digital Technologies Digital Technology Educational Psychologist Energy Management Health Insurance Health Maintenance Organizations Hospitality & Construction Human Centered Design Integrated Decision Approach Integrated Decision Making Intelligent Document Processing Kaizen Medicare Moodset for Excellence Natural Language Processing (NLP) Offering Managers Oil and Gas Optical Character Recognition (OCR) Pharmaceuticals and Life Sciences Photographing Price and Routing Tracking (PART) Process Design Document (PDD) Product Identifier Descriptions (PIDs) Python Quote to Cash (Q2C) Resilience SAP Sales Quota Team Work Telecommunications Text Mining Visually Displayed Work Culture master text analytics virtual resource management