 Hi all, welcome for the session. So this talk is going to cover about AI and how AI can help to the human in the society in a much more effective manner having novel innovation features and how a research can be converted towards a commercialization and I would say monetization. So this is Sachin Agriwal. I carry the experience about 20 years in the industry having research and innovation conversion towards a commercial or the market ready products or the projects I would say and considering the various business aspects from the servicing point of view also. I carry AI PhD degree along with the lots of papers, research papers and patents, almost 25 patents, 25 research papers, five book chapters which all covering about research innovation in AI. So in this talk today we are going to talk about how the solutions or ideas can be converted from the raw research to make it more business effective which can be used for a particular segments or the technology. So here I would say starting from the five pointers so how the research can be taken up for the monetization. So in this case let's go one by one and the later section of the talk we will be talking more about how the business or the particular segments or a particular problem statement should be selected to have the research development with the support of AI. So first of all let's see how the domain selection should work. So first of all idea should be predefined in a boundary of a domain. So that's the reason we need to select a particular domain and the second point is in a domain we need to really understand that this domain is suffering what type of problem or I would say issues so that we should come up with the right solution. From a point of view for the business or the research we should be really knowing the pain points of the business and reasoning of the particular domain that's why I suggested is to focus more on the domain and if you select a domain and identify the need or the pain points of the business then only we should be moving towards the I would say solution. Solution should be definitely the third point is adding the more value over here is we should be having the solution or the novel approach in a much more effective manner and much more I would say differentiating factor thus should add the more value what current technology or methods are not able to provide and definitely that's the reason we need to move towards the market's acceptance that's the reason it will be the highly monetized value. So the just to summarize again so solution should be much more solving the business need or the pain points so that the person or the service which is going to use it out make it more effective solution oriented and that's how it's a win-win strategy. So let's I would say the move for the next one and just see how the further AI and Genia can further support. So Genia is basically like an emerging trend which helps to the users in a much more effective manner. It just also converting the capability to generate the responses as per the need. So it's not a very fixed predefined I would say a template or set of classified output. It's more effective more generative more creative and a more personalized manner so that once as per the more personalized solution we can connect more targeted users in a business and that's the beauty of the Genia solutions. What the Genia capability so it can analyze the user content it can consider to create a specific content example the social media post or the newsletter. So currently if you see the content of the newsletter or the media is published for a certain group of people or a society or a classified manner but what if if Genia can deliver to me the things in a much more effective or a personalized manner. So it would be more about the engaging the customer engaging the need and delivering the solutions accordingly and definitely it will increase the revenue. So how this process will journey of the monetization with Genia because once we target the more personalized way it would be the more effective way of reaching to the more I would say even in the beginning I suggested that more user need or problem oriented. So the reason is if we reach it out the targeted user targeted problem targeted solutions are there so Genia can support it out. So that means we need to create a customized solution. So in this way we have the more variety of the I would say the data or the content and generating the output accordingly and it would be more on the subscription to one or more on the user driven on the need. So Genia is I would say another helping hand as a technology people or the technology experts or the developers. So we could generate the more customized solution and Genia can take that kind of load. And as a human we can add more value or a mere I would say as a technical experts we can add more value by adding the Genia capabilities. So what's the basically important is with Genia because Genia is really a powerful tool in the current world but it should be responsible right because once we develop a solution or the technology as a techie guy we need to really need to see that how Genia should work in a much more responsible or the balanced manner. So that there should be a proper trust by the data or I would say society output what we are delivering. And it will be I would say a hand shaking in a long term scenario. So definitely Genia usage to be controlled, responsible governance driven should be there so that we can leverage the Genia capability in the right and the positive manner. So it is I would say our control should be in much more effective manner. And I would suggest that that should be also part of our project or the research when we prepare. So AI is basically meant for the right information at right time for the right user. So that is the purpose of AI. So that if somebody needs X information at X point of time for X purpose so that should be delivered in much more a right manner and that's the AI comes into the role and that's the reason that Genia is going to do it out so that if I have the 10 users in my business to solve the problem so that 10 users may have the different expectation at different point of time in a day. So that's why AI is going to help it us. Another thing which really I would say it comes up in the research and news and media cell that AI versus human intelligence. So in this case like human is human right and human carries a human intelligence which is far beyond the AI. So definitely human is controlling AI and if we do a right governance, right control definitely human is going to control the AI. Definitely it's a part of the consideration and AI cannot replace the human intelligence. So in this case definitely AI will be the develop the part or will be development part under the control boundary I would say. So for example if we consider the banking example AI is going to AI we cannot replace an expert. AI is going to helping hand to the particular SME to do a more I would say the task in much more effective manner, more higher effective manner and the more number of things. And I would say more data oriented decision or the I would say the process to be taken care by the SMEs or the person or the expert who is doing the his job. So that is the reason AI is going to just one of the helping hand for the SMEs. Let's take an example like how this AI can help it out for a particular segment and how we should basically generate a problem statement come up with a solution and deliver a best solution which can lead towards the commercialization and the monetization. So natural disaster let's look it out the situation. So here like for example like what are the different type of disasters? I select like I select a domain like I have to work for AI project with natural disasters. So here it looks like the different type of natural disasters are there like wildfire, earthquake, flood, drought or they could be more number of I would say disaster which as a human we cannot control but as a human if we get aware timely we can definitely get the certain I would say reduction in the loss, more saving of the human lives that could be done using AI. So if we know, so in this case when I look at the scenario of a particular disaster problem so it can have the two type of scenario how AI can help basically to control such kind of I would say savage of the society or I would say the resources which are particular going to be impacted by the natural disaster. So this is a one area by understanding through AI what is the intensity of the disaster, what's the time, how long duration is going to be there. If we know that thing using AI so this is one section I would say if as a solution I would say this one area can be considered for AI projects to be considered. So look it out in the scenario as a one scope and consider how AI can deliver solutions by considering the data oriented approach and look it out what the best effective way to save the human life or the resources by considering AI. Let's take another example of the banking let's say. So for example in a particular area it's going to be disaster let's say in terms of some flood or et cetera. So let's take an example is a flood area on a coastal area or somewhere so it's let's say disaster is going to come. So as in banking like what we can do it as a bank what should be the things that bank should support. So for example there is an A bank which is going to be considered what should be the precaution project opportunity for us. So let's divide and conquer situation. We divide the situation in three stages. What should be the precaution measurement before disaster during disaster and third is after disaster. So these are the three points that we can take into the consideration. So before that I would say the disaster what should be the precaution so as a bank that I should be prepared and ready hey if the disaster comes how would I going to save the data or maybe how are going to save my assets whether it's like a digital or the physical assets that's the one preparation the planning that should be considered by the banks I would say and in terms of data so definitely they should have some data migration policy data migration patterns or standards so that data should be procured and safe always. So this would be like a pre-planning. Second also would be like pre-planning would be considering what type of loans would be there in the pre-stage if we have to give it out. So according to the premium I would say interest or ROI return on investment should be planned in that way. Now the second point comes in a caution for the AI driven project if we really need to see hey as an AI what could be the impact on the disaster in the banking services and in particularly let's say for the how much area it's going to impact 50 kilometers, 100 kilometers something like a range driven. So that could be in a planning stage which AI can do it out. Another important thing considering example of the bank like for example during this disaster the situation would be there like how people would be bringing it out their money back from the bank and at the time it would be quite important. So bank really has to really okay what would be type of the deployment of the services during the disaster time. So this is the one thing which becomes quite important as a part of the planning. So in this case as a bank and plan AI driven feature say hey what kind of different I would say expected I would say outcomes or expected demand of the money which would be coming it out as a digital currency or I would say the currency of the physical mode. So that could be part of the during disaster and secondly without how much seamlessly this banking services can be started or resume again during the disaster time and third again after the disaster let's say for the banking example again after disaster people may come up to the bank again to demand or ask for the more loan. So in that case what should be the bank's approach how can be more effective from the business oriented and the solving again the pain point of my banking customers which are associated with me or maybe the new bank customers may arrive to me. So these are three strategy or the three different categories are there which can be much more effectively planned. So if we look it out again the more detailed version. So loan would be there what are the risk mitigation approaches are there. So during or I would say before and after would be there and I would just suggest as a second time that the money giving or in because sometimes ATM may not work at that particular time. So in this case that situation will arise again. So during disaster every bank would be delivering or be trying best to do it out but how if AI can be helping us as an helping hand in a pre-planned manner. So for example like bank site maybe shut down due to the disaster possibility but in this case if bank has a pre-planned they may have some I would say the mobile banking facility or the bank on a field. So for example bank would be coming on a truck or maybe a van or some car, et cetera or the vehicle. So in that case as an AI now we have an opportunity over here so like one van or a car or a truck can cover how many villages I would say or how much kilometer range or particular area. So AI can do a such kind of predatory classification. So this van can accommodate let's say five different villages or five different sectors and based on this sectors how much is the demand is going to come. So how much cash a van should carry in a particular day. So that could be done based on our data analytics. So here now role comes on the AI. It will be doing a classified data approach so that it could be making a much more effective I would say prediction so that when the van goes from the out of the hit quarter or some let's say regional branch. So it carries sufficient amount of the I would say cash or the rupees. So that it should be a seamless experience so the van goes in a particular area. So sufficient cash is there. So AI can help it out that particular way. So in that case we can estimate how many users are there, how many users are residing in a particular area so that the bank van can fulfill the right requirement and can coordinate and so that bank can also plan how many vans or the vehicles are required to be deployed in a particular area. So this is something like from the point of view how the cash or amount should be distributed during the disaster time. So now the interesting factor is another situation when you look it out one more important things will come it out about the verification because people might not be able to carry for example there I would say checkbook or maybe I would say another way of let's say ATM card or credit card or debit card like that. So those things as a physical one may not be applicable or available. So how AI or technology can help it out over here. So that's the reason the bank has to prepare a bit early in a stage and as a techie guy we need to really support what are the different authentication method would be there because at the time as I mentioned about the bank van will be there to deliver the amount the user or my consumer or customer will be there but he does not carry any checkbook or credit card, debit card. So now the situation becomes really tough over there. So now the thing is important becomes role about the biometric authentication of the user. So at the time the bank also has to consider what type of different authentication method has to be considered. So here the role comes about the biometric authentication where it's like a fingerprint or retina. So bank has to consider and a pre-plan in an early stage so that the right person gets the right amount of money at right time that's why we're linking it out. The person or the AI should be able to help to the person in a right point of time by the right mean and a right method at right place. So that is the beauty of AI it's going to help it out. So even this scenario as we try to consider or I would say convey about the methods or the messages or the tips for how to prepare a project and how to be more I would say pain point oriented. So the solution should be really solving a pain point or a particular user or the business. So this was you're talking about the disaster. Now we will be talking more about slightly that how AI can help it out for the banking customer facing services. So in the bank when we say there are lots of services which can be supplied or serviced by the AI. So for example, fraud detection, data migration, data saving and they could be n number of things like fraud detection for the credit cards, et cetera. But in this talk I'm going to talk about more about the customer facing. So where the bank and the consumer interfaces and this example I'm taking an example but can be applied for the various other sectors also where the business basically interacts with the consumer and that's how the certain tips by understanding the consumer behavior, consumer pattern we can really come up with the right effective approach. So let's look it out what are the bank objectives. So in this case let's consider the different AI approach how it can help the bank I would say. So it can enhance the customer experience, the way we interact, it can innovate and I would say compare and compensate to the right thing. And it can reduce overall operating costs and definitely can do the revenue. And finally it can take the risk control I would say. Let's take a very simple example. Let's say the one particular user he really needs a loan for a home and he really needs to understand what is home loan approach and how should we go for the approaching and applying for the loan. But this person is not aware about the process overall. So firstly he might be going to the bank or a bank representative to understand hey what is the loan and I just need to apply for the home loan for a particular sector and he may have let's say the several queries example 50 queries on day one. But when he goes for the second bank he understands the process but he has the query that how to compare from A bank to the B bank in terms of the policy or the process on the pros and the cons. So now what we see the pattern he may go again to the third bank again and might be possible he would be coming back again to the I would say the first bank again to see what is good ban and how it can make the effective output or benefit for the person. So here the situation is the person is trying to grab the information for the loan related from the various bank understanding the pattern, understanding the issues he may face and the planning. So what we see as an issue just to take a loan for home he's trying to struggle and spend a big amount of time to understand the banking process. Now what if AI comes and help the bank as well as my customer to understand hey this is a loan and how the best way to look it out. So in this case like being an AI and the banking side if bank offers a process and AI driven features so that the bank I would say the person or expert or the banking officer who is the part of this conversation he can make the more effective communication and start from the right query understanding the context of my user. So that's why the AI can play a role to understand my consumer and solving the right question answers pattern at the same time. And here definitely JNI could also do a helping hand over here to derive the question answers in a 24-7 in a much more effective manner. But yes again as I said about the role of the banking expert or SMB remains with him to decide whether to give a loan or not or maybe if to give a loan what terms and conditions should be applied. So that is still again lies with us as a human as intelligence to take a decision. But how AI is going to help it out. AI is going to help the more data oriented approach where the banking expert would be sitting with the more data what's the background of my customer. How is capable and expert to do it now or maybe to the premium in the coming years. Similarly my consumer also understand here this bank is going to deliver this, this things. This is a pattern and I have this type of conversation whether it's a very day one conversation or day three conversation. So both party understands what's the need and the requirement and AI can be a helping hand I would say at that particular time. So this is how AI can help both the parties and very basic structure when we look it out. It's look it out about the data and experiencing or maybe extracting based on the experience of this feature data set and doing I would say classification making them more effective created I would say prediction at survey. So in this is a very basic system for a banking and the customer when I look for the interface side and it's how it can be win-win situation for both the parties. So when we look it out whether the different trips or the trips can be further there for the I would say projects planning in AI driven banking features. So it could be again go for the product roadmap planning. For example, if person is let's say today taking home loan tomorrow he may come again for let's say for the car loan or maybe some of his business loans. So what could be the different products we can plan it out. So based on the strength and wealth of the particular person, yes definitely we really need to treat the data in a much more confidential manner. Yes without much impacting the personal data I would say and that should be under the boundary out of it. So that product planning or the roadmap can be decided based on such kind of pattern. So we can look it out for the consumer side what could be the next potential features or the projects or the products that we can do it out. And definitely again it's a price sensitive because customer should not be going or knocking the door of other banks I would say so in that case it is quite price sensitive. So whatever the services are offered that should also be considered what we are offering in a bookie of services along with the pricing strategy. And definitely all the tuning of all this modeling point of view and the feature point of view it's really required. And definitely we really need to understand the I would say sentiments or the behavior of the consumer. For example I would say let's say he's planning for a particular home loan in the month of let's say January 2023 but let's say after three months or four months the market can impact it. It really impacts the industry. It really impacts the market. So in that case what as a bank can help to a customer or what customer might be having such query or situation to change his I would say the loan related I would say the premium or interest and et cetera. So it's just like the both side that AI can also again help it out. So it's a both way AI is going to help to understand more of the consumer issues which may be direct or indirect I would say. And similarly from the banking side also like bank may be rolling out the more new services and he really wants that whatever the customers or consumers are there they should be associated more. So here bank can also use one strategy let's say I have like 10 users or 10 customers associated with my bank whether in the future I should retain these 10 and I should bring the next 20 set of the consumers at my side. So business and the bank definitely will be looking to associate. So that's the reason AI can play a role over here. So now looking more about the GNI and the banking so definitely GNI can also help to the banking sector. So even it's also seen for the Gartner's report that it predicts that 20% of all of the test data on the consumer facing use case will be definitely going to I would say increase by 2025 and GNI could be supporting on a more digital world. The more application towards the gain oriented and definitely there's a potential for more personal data oriented and in this case GNI can really help it out from handling from the unstructured data as well as the structured data. And it's more like GNI can learn the pattern from the generated I would say the algorithms and LNM models and look it out the strategy to make it more effective and looking it out the more I would say apportion T having a synthetic database learning so that we can have the more effective features in the sectors. So it would be more in tracking I would say again the I would say this going to add more value for the automation efficiency and the instant responses. So definitely AI is going to available to the user as per the needs. So in that case the person does not need to wait for a particular I would say the service provider. So he can ask the questions. He can prepare I would say responses accordingly and it's going to run in a parallel experience. Now if you look it out what are the different business operations which would be there in the banking sector. So if you look it out for the person is going to open a new account. He may have some queries. So in this case GNI can help it out to do a let's say efficient way of KYC by having the GNI based approach. Enhance banking services so that our banking can provide very customized personalized I would say SMS to the particular person. And in that case along with this virtual or the digital word experience I would say. Secondly again as example we discussed about the loan side. The loan could be again for the small business or the larger business. So accordingly it could be like very customized way of handling the loans applications or the loan queries. And again for the commercial I would say banking whether it's a business for a I would say applying for a small industry or a medium industry then it could be again. And definitely investing banking. The person may come to the bank also to do an investment from the mutual funds or some policies or et cetera. So again this is going to help it us in a very much more customized manner because we're really going with the technology. We're really going on our products and the services. So again it's meeting the right expectation of the consumer with the right services. Now if we look it out about the different financial goals and the preferences we discussed about like person may have his own. Let's say he has done for the I would say the loan but he has to some do a goal for some planning for saving or maybe some his investment on getting some investment return. So this could be another way of a bank and support some services about the saving account or the big purchase which user can plan. For example big purchase like purchasing a car or vehicle. Again Jenny I can do a basically prediction for the need of my consumer. So in this case it may go for approach. Let's say he's planning for certain trip or he's planning for some travel. So those kind of things which will be again a part of Jenny I and it can understand the pattern and maybe suggest the right amount or without accountability I would say. And in this case it would be a right connection based on the Jenny I features understanding basically the customer oriented I would. And similarly again in this case it learns the behavior of a consumer when it being connected from the healthcare point of view or from the well-being point of view. So in this case if there's some product or services which can help to my consumer for his wealth or I would say again with the health side also. So in this case this is a combination for the particular consumer for him or her as per the family members. So now even you sometimes see there are lots of services offered by the banking. They don't go for a logical fashion. They come on randomly fashion to the person. So here in this case what we are suggesting as a new emerging I would say solution could be there. In this way we go very structured way of as I discussed and explained about the logical fashion of the need of my consumer because he may have the X policy today possible Y and Z. It has to go in logical sequence. So it should not be Bombardier is a parallel one. It has to go in logical fashion. So my user can select one thing at a time as per his pattern. Might be possible the other consumer in the pocket he needs the three services at a time or even the parallel processing. So in this case that would be a very that parallel processing approach. But here is like a very sequential time driven approach or maybe the his or her preferences approach. So that's why the J&I can take care of. So in this case again the J&I would be having some challenges of the risk about the security and the control and the data privacy again is a prior. I would say that one of the concern. So that AI has to take care along with the SMEs experts. So these all a big combination I would say it has to run as ecosystem. It has to run in a basically a closed environment so that we should not have any hindrance by the certain other things. And it should run in a more controlled and a sequentially other secure governance oriented model I would suggest. So AI can add more value to the expert solution. However it cannot replace a human intelligence. As we discussed in the example previously like banking expert or the role was there quite important for the consumer. Because consumer was coming for the loan or the policy but final decision has to be take place by the banking office or the expert. Now with AI what is going to happen? This person or the banking person is going to be more data oriented. For example if customer comes to him and ask hi why my data or my I would say the loan application has been rejected. So this person can be more data oriented, more factful and they can explain also. The role of expert becomes there but it would be suggested with the more options or the more data facility. Now if we look it out the more I would say customer driven approaches for the insurance because bank also deals for the insurance related features. Now for the insurance for the banking side let's take an example of real life. So what happens if the car is damaged on the road and currently without the AI the person or the driver goes to the workshop or I would say workshop going it's a bit tedious process first of all. Then reaching there getting appointment for the banking officer or the insurance person to come and approve and do an investigation with the right amount of time. Then only the repairing work starts. But what if AI technology can help us over here? So the person who owns this vehicle makes a video with the damages and send the instantly to the right insurance party as per his insurance policy. An insurance person looks out the video and in this AI can help it out to bring it out that features which really require attention for the approval from the insurance point of view. And then he approves with the customized let's say the policy based approval and send it back to the workshop and it's process basically process to do a I would say rectification or work by the workshop. So in this case lots of paperwork got resolved and it's instant approval and immediate processing. But here the role again becomes important the person who is approving the policy but it's make very simplistic easy experience for the consumer who is dealing with such kind of damage over the time. So this is how that AI can really help it out and solve the pain points of my consumer. So this is how another example that we are solving the pain point of my consumer. Let's look it out the different customer services when we talk it about. So in this case in this customer service let's take an example what data is trying to be considered for the prediction. So in this interaction we'll try to look it out what is the health behavior, things which are maybe discussed with the person and definitely the privacy is a concern that has to be taken care. And in this case we can look it out the classification the history pattern of the my customer. So in this case what are the different services can be offered. So in this case the service can be offered by the verbally or the display. For example display could be some nearby or display could be some push method in an email message or some other media. So that right information can be pushed to the user and not as a bulky or I would say a junk kind of information. And so that the information is really searchable and based on the data and the facts is there. So that finally the decision can be made meaningful accurate and based on the more data and the fact oriented I would say. And finally it's going to help the bank as well as it will I would say boost the overall customer relationship and make the promising hence on experience for the long term association. Because if bank is going to target the right information at the right time so it's going to be a long term association. So this is how this customer service can be considered in a text image. For example text can be SMS or text could be some message it. Verbal or I would say the audio form acoustic. And definitely if it has to be visual it could be image or video. So this is how it's going to work it out in the customer experience. Now let's take a simple example of healthcare in this talk we try to cover it up in a healthcare in a shot. So for example digital also plays another role with AI how to deal a much more effective manner where we have something things in real world a certain parameters let's say X features and we have the digital world which can simulate those X features. So this thing can be applied as a digital twin concept. We can do a simulation in the digital world in a much more first effective manner and we can do a prediction of whatever variations could be faced when we go for the digital world simulation using digital twin. Let's take an example the similar one. Let's say the real world has been considered and we are moving towards the digital world and we see the five, six variations in the digital world simulation depending on the future prediction policies I would say process or the situation. But in this case the person or the experts he considered hey green color is one of the possible digital world simulation and this would be the more possible one. So the person or SMEs can accordingly take a decision in this way. So similarly for the healthcare the medical experts or the doctor who's going to do a simulation with the digital twin so that he knows if I apply this type of treatment what could be the output would be there possibly but if we do a simulation in the digital world or digital twin so he can consider here this is the best simulation for my patient and I can apply this type of treatment so that this treatment looks suitable and the best and again the doctor decision remains the same because he really is just going to use AI but AI or digital twin is not going to decide. So overall AI remains as a helping hand to the doctor to decide what the digital pattern or the digital twin output comes out. So in this case this can be applied for various scenario I would say for the medicine generation and data oriented approach for the healthcare records I would say and progression modeling and I would also suggest like virtual patient simulation and the chronic disease which is something like a disease like if it is considered or maybe identified detected at the right time so that doctor can start the treatment accordingly and through AI if we are going to track the diagnosis I would say in this case definitely is going to be a much more effective I would say simulation oriented approach for the doctor so that doctor can save more time and maybe come up with the effective quick solution using AI and definitely this healthcare side AI can again be used for the enhanced medical imaging for rehabilitation of the therapy, drug delivery, disease reassessment I would say virtual health and a remote patient monitoring. So this is how it's going to help it out in a healthcare scenario. Overall look it out that AI can also help for the various application it could be for example 60 also. So in this talk I discovered a very short part where the digital twin can also help it out to understand the network behavior or the wireless data transfer. So we can apply the learning from the virtual world or digital world using digital twin to the real world what type of scenarios for the effective data traffic management or maybe demand of the traffic for example video data or maybe the 3D data kind of transmission. So in this case we can look it out what are the network optimization what are the prediction techniques for the data and the channel capacity and how can we make it much more effective data rate communication having the smart antenna array system. So this is also another 60 can also leverage with AI and digital twin. So this is how the various example were considered over here and finally just short to summarize. So as we discussed about the lots of tricks and the trips according to the domains and the applications like for example the right information should reach to the right user at right place and time and considering the different industrial application and the research. So this is how you can make it much more effective your research and the project and make it executable at the right point of view according to the demand of the industry or the services I would say. So this you can use it out as a tips and the tricks and make it more effective way and try to leverage in your projects to make it much more effective impacted manner presentation or deliverables of your project. I would suggest to use this session much more effectively and apply it to solution and special thanks to AICT to make it possible and making reaching and to you and again lots of thanks to all of the viewers and the listeners for the session. Have a nice day, take care.