Digital Transformation is a Continuous Spark: Vishwanath Ramaswamy, IBM Systems

Several things are transforming IBM in all the new areas, be it digital, blockchain, or Artificial Intelligence; and this is also helping clients transform to the digital era

In an exclusive chat with Ibrahim Ahmad, Group Editor, Dataquest, Vishwanath Ramaswamy, Director, IBM Systems (India& South Asia) talks about Artificial Intelligence and the new goalpost in the digital transformation world.

With everybody talking of the Digital Enterprise, what’s the big news at IBM?

A lot of things are really exciting from that perspective at IBM globally as well as in India. Several things are transforming IBM in all the new areas, be it digital, blockchain, or Artificial Intelligence. A lot of these are really converging and we are helping clients to transform into the digital era, and in the process get clients to move into the AI era because that is the strength of IBM. If you see our history, our real strength is in terms of encompassing a lot of software services and solutions management, and taking the clients on a journey of on premise to a hybrid cloud or hybrid cloud to a multi cloud or from a multi cloud to AI. There are different journeys which clients are going through in this disruption era. And now, primarily it is about taking enterprises to an AI journey. You would have also heard about our Red Hat acquisition. That’s adding to that excitement because the Red Hat acquisition is complimenting to the strengths of IBM. We would be the dominant player in the multi cloud and the hybrid cloud era.

So you are implying that AI is also part of Digital Transformation journey?

If you look at any enterprise journey and their evolution cycle, digital transformation is a very vast area. Mobile banking apps on your phone are also digital, making a transaction using Paytm is also digital. But these are very nascent forms of digital. The question is what after that? And how can an enterprise move further ahead in the chain by leveraging these new technologies to give a great customer experience; and also get some insights out of the customer information and experience; and then to suggest even more relevant products or services back to them. Digital transformation is basically an entire journey and the organization could be at a different stage, but it would culminate into an AI system. With the AI the inference our clients can improvise their own systems to again drive more business, so it’s always an infinite loop.

Is AI therefore the end point in this apparently never ending journey?

No, transformation doesn’t mean that there is an end stage, because transformation is always going to change and change is not constant. Disruption is always going to be a new normal. Disruption could be in different forms and states but in today’s era, AI is the new disruption. But you never know, and in 5 years AI could be the starting point, and deep learning could be the new horizon. Or quantum computing could be the next horizon. So with digital transformation, it is a continuous spark that organizations will get into. Today at IBM we see quantum computing to be the next horizon doing really deep learning models for solving those problems of businesses which our classical computer cannot solve. That’s a transformational journey and the end state is not AI. The horizon that is visible is AI and quantum computing.

How have you have made the digital transformation journey more tangible for your clients?

There is no end point to this journey and it is an infinite loop. Organizations have to decide what they want to do and what is the time span they have to sit in the train to go to the next journey. Some clients say they have everything done but there are too many manual interventions so how can I automate it. That is one journey. There are others who are in the advanced form of the digital transformation journey so they have started AI in one or two places. Now how can they make AI pervasive in their entire organization, and that is another transformation journey.

Are there use cases of AI in only some verticals or all across?

Absolutely and undoubtedly, all across. I cannot name clients but I can give you a lot of global clients that are referenceable. We have large Indian clients in the healthcare, general insurance, and banking industries. The bank is doing cheque truncation using AI. We also have another couple of other POCs going around in other banks. The Government of India also realizes the importance of AI as they have sanctioned a special budget for a national AI Centre. Growth and spread of AI is rapid and pervasive and we strongly believe that tomorrow organizations won’t say that this application won’t run on AI. The question will be, is this application AI, or that if this application is AI ready? I can’t share the numbers unfortunately but hundred percent of the clients are on the digital transformation journey. They are in different stages of the journey; some are in the advanced stages, some in the nascent stages but they are all embarked on the digital transformation journey. Segments that are very active include banking, healthcare, R&D institutes, startups, telecom, and manufacturing. In the government sector also there is a thought leadership for AI that is going around and it will get there sooner or later. But other segments are also taking cognizance of AI.

Are there some statistics available with IBM to indicate how much productivity improvement has been achieved?

You can’t really put a number on it; it could be 2x or even 10x and so on. Improvement for one bank could be 2x and for another bank could be 10x because they possibly may have automated to such an extent that the benefits they see may be different; also the amount of data they have to make the AI model successful is also a big dependency. The more data you provide the more accurate the business outcome.

The AI phase of digital transformation is fully dependent of capturing partner and customer data, feeding it into AI systems and processes so that it can be used as timely, intelligent and actionable information for enhancing revenues, creating new business streams, improving customer satisfaction etc. However, there is a big lack of best practices when it comes to data capturing, management and protection in India. But how prepared is the country when it comes to data.

I beg to differ on that. Data for any bank or any organization as a matter of fact is available. There are four stages of an AI. The first is ingestion of the data. Second, they work on preparing the data; 80% of the time is being spent by organizations on the preparation of data and for preparing the data is not just the data scientists that are important, there are domain data scientists who have knowledge of different domains of the industry like insurance, banking, etc. In stage three, organizations want a business outcome which they have set for themselves, and what data will they use to code a model or write an algorithm and AI algorithm to train the machine to give that inference. So the training part and inference part which is the stage 3 and stage 4 takes only 20% of the time; 80% of the time goes into preparing the data. The existing data within an enterprise today and the probability of accuracy using AI is very high. Organizations, kind of, train the model in such a way that the accuracy steps towards 100%.

What kind of increase in RoI do you suggest to your customers?

With all our clients we do a POC (proof of concept) and a pilot; and compare it with the existing way they are doing it. For instance, there could be a cheque truncation system which a bank is using that delivers in 24 hours using AI instead of the 48 hours that it took using the traditional IT solution. This is a 100 jump in productivity. It might not have a direct revenue and business outcome but it can really help in productivity outcome. Today your target is one to many but when AI and deep learning come into play, it is either one to one or one to few. Then the client experience becomes very different.

In organizations today, which CXO is most excited with AI and the solutions it offers?

The business guys. But it also depends on the use case. For example, credit card product recommendations in a bank the person who runs the credit card business for a bank is the most excited. When it comes to AI use case for regulatory stuff, which enterprises have to do with their regulators, then the CFO is most excited. If we are doing an AI algorithm to go through all the legal documents then the law and attorney officers are excited. It’s in fact not just business people, even HR managers are excited about AI. I had a conversation with someone from HR who said it would be interesting to predict for how much time this person will stick with the company. One HR client said that the number of people joining after the offer is given is shrinking, so they asked us if we can build an AI which is able to predict the probability of the number of people joining after getting an offer. Because the youngsters and millennials post it on Instagram and Facebook the minute they are made an offer, so based on the tonality and mood of the posts, the system can predict if the individual will take the job. That’s the true AI.

AI systems can even be misused. Are there some mechanisms to prevent misuse of this technology?

Data protection and data privacy will be very critical here. Data privacy is a two way street, which is as an individual and as an enterprise. We need to decide how much information we put out. The choice of an individual to be on social media is purely his or her choice and an individual decision. The minute one has decided to be there, there are enough tools that prevent external forces to intrude into your space.

Is AI the main thing that IBM is putting as goal post for customers?

No there are a lot of other things too. AI is only the horizon to it, but what comes along with it is when you look at 4 stages of building to an AI journey. Data is the most important, so obviously how secure can I keep this data is also important. Then how can I make it agile for the client to run the data and take an inference it’s a strength of ours. And how do I smartly optimize the service. And where do I run the data – on premise as a private cloud or on public cloud. As I said at the beginning, a Digital Transformation journey has all these phases and IBM’s strength lies in all these is phases. Then only can the full journey be delightful.

What does the IBM ecosystem involve when it comes to working on AI solutions for your digital enterprise customers?

The ecosystem involves two parts, one is our value added partners and large SI integrators, the other part that is rampantly growing is our startups, the developer ecosystem in the AI space. Focusing on AI we have about 20 partners. We also have our own services. So it’s a combination of all of these. We have a huge developer ecosystem community in India that uses our APIs. So you don’t need to start from ground as you get a headstart with us. Use APIs which is your domain of expertise. And this ecosystem is growing by the year. They just focus on the mission and the outcome and we help them in this journey.

Leave a Reply

Your email address will not be published. Required fields are marked *