IBM ushering local corporates towards AI
By Uwin Lugoda
The International Business Machines Corporation, better known as IBM, has been a big part of Sri Lanka’s tech ecosystem for around 57 years, and during that time, has made a significant contribution to the nation’s technological progress. This was showcased in 1969, seven years after coming to the country, when former IBMer Tony Kandaiya set up what is believed to be Asia’s first IBM 360 in Sri Lanka.
Similar to Kandaiya, IBM has looked at Sri Lanka as a country that is willing to adapt and try new technologies, allowing for IBM to position their technologies here. IBM currently has around 15 to 20 partners who actively work with the company, creating a good partner ecosystem. These include the likes of Blue Chip, MIT, Enable, Cedi Technologies, VSIS, and the most recent being Fortuna Global; each of which specialise in specific areas. It also serves a prestigious set of clients from several economically important industries in the country.
Last week, The Sunday Morning Business sat down with IBM Sri Lanka Country General Manager Riza Wadood to talk about the company’s focus on the country, how it builds its clients, and the Sri Lankan tech ecosystem.
Cloud is a big focus area for IBM. In your experience, how willing are Sri Lanka’s businesses to migrate to cloud?
I think that the only reluctance we are seeing today is regulatory. The customers are willing to migrate to cloud, but they are basically constrained; if you take the telco or banking industry, the Central Bank of Sri Lanka still doesn’t clearly state how the data needs to be managed, because they don’t want the customer data to be in any other country but Sri Lanka.
But where it is not regulated, most of the customers have gone to cloud. ERPs (Enterprise Resource Planning) are running on cloud today. Whoever did on-premises implementations in Sri Lanka five years ago, are now, when approaching renewal, want a cloud proposal as well.
They look at the cost effectiveness and most are going towards cloud because it is subscription based. You don’t need capital expenditure, and you can also keep increasing your resources on a needs basis. When you buy hardware on-premises, you need to size it for at least three to five years but buy it immediately, leaving that hardware idle for some time. When you go to cloud, you can buy what you want today and one year ahead, you can keep increasing it step by step.
Cost wise, there are certain nuances, but they appreciate that in the long run it has more pros than cons.
How much guidance does IBM offer its customers in the cloud migration process?
We are now a fully fledged cloud partner for our clients.
It starts from advising somebody on how to go to cloud, and once we have, then it’s about moving to cloud, after that, building and managing on cloud.
When it comes to advice, we are cloud agnostic, which means we advise a client irrespective of whether they want to go to Azure, Google, or any other vendor. So we are a technology partner.
Then, when they want to move, we will help them move to any cloud, and when they want to build on cloud, we can build on any cloud. When we manage, we have something called the “Multi Cloud Manager”, where you can manage multiple clouds from a single console.
In Sri Lanka, while being the principal, we don’t just sell the cloud; we assist with the advice, the move, the build, and the management. Not many other principals would provide it end-to-end like that.
How has cloud adoption been in Sri Lanka so far?
I would say pretty good. Some of the smaller companies are very keen to move to cloud because there are so many advantages. What cloud brings to organisations is agility, which allows you to innovate faster.
For example, Dialog has APIs that start-ups can consume and develop applications on. So similarly, when you’re in that context, it becomes so much more agile for you to operate on a cloud framework.
Not only do you have what you need, you can also start consuming so many other services that are available on the cloud, like Google Maps.
How were companies like Uber born? Google came up with Google Maps, and they started exposing APIs people could start using. That was their income stream. Now, not only Uber, but PickMe and so many other companies, are using it and it is driving innovation.
Similarly, with telecommunication companies, when they start opening up these APIs, a lot more things become possible. Things that they don’t want to do themselves, others may want to do. So we have done something similar.
You can’t replicate the IBM Watson Platform, our cognitive computer, in every country. But if you host it on cloud and give it as a service, anybody will be willing to pay 25 cents to consume that service. They don’t need a million dollars to use it and even a student won’t mind experimenting with it. So they can do their own research; they don’t need to think about wanting their own hardware, and they can develop their own products by leveraging those services.
So we have about 50-odd services that are now provided free of charge on that platform for people to consume, whether it’s speech to text, text to speech, personality analyser, tone analyser, or something else.
How focused are you on security, from both a cloud perspective and in general?
That is one of our main focus areas. To put it in order, it’s cloud, AI, security, analytics, IoT, and blockchain. That’s basically the Fourth Industrial Revolution, which is where our focus is, and security is one of its key pillars.
If you look at the Gartner Quadrant, the product that we have is positioned right on top, and now what we have done is built on top of it and have something called cognitive security. Quite a few banks in Sri Lanka are using it. What they have is a platform that runs within the organisation, which we call QRadar. What it does is it helps organisations monitor all the activity in their network, whether email, internet traffic, or the people accessing the systems. All of this can be monitored because it is an event in a network and it is logged.
So all these logs can be analysed by this tool and it will start throwing alerts if it sees any anomalies. In this instance, an analyst needs to look at those alerts and see if any of those are malicious or if there are any issues with that, after which they can see how to block those threats. Since those alerts are sometimes high depending on volume, you can send those alerts to our cognitive security. That AI engine will analyse those alerts and tell you that you should only worry about five or six out of those 100 alerts.
We call it Watson for security and it is a huge thing. We are also working with some of the stakeholders in the country to set up a security operation centre (SOC), because Sri Lanka doesn’t have one. What most people have is the tools within the organisation and those tools are managed by the internal teams which may not include security experts.
If we have something like a SOC setup here, then these companies can subscribe to this SOC, and it will basically monitor this network for the alerts that are coming and tell this security team sitting in these companies where the vulnerable parts are. It could be malware, viruses, or malicious emails, and from what we see from one client, could be around 10 attempts a day.
So it’s a continuous effort. Luckily, Sri Lanka might not be such a vulnerable country compared to the bigger economies. In places such as Hong Kong or Russia, these activities are beyond what we can imagine.
Can you speak about some of your biggest clients in Sri Lanka?
We have the state banks along with many of the big private banks.
Has IBM tried to collaborate with the Government on the digital economic drive of Sri Lanka?
We have not directly got involved, but if you take most of the large projects, it is handled by our partners.
They may use our hardware, our software, or our services in some way, but we may not be directly involved. We aren’t the only ones however; there are others involved as well.
Even in some of the large projects, we are involved from a concept perspective and by bringing down our thought leaders to Sri Lanka. Because we may not have all the components, we may provide some of it and the partner may bring the other components and do the needful.
Can you highlight a case study of a company that has deployed IBM solutions with successful results?
Hayleys Group was a public case study; we implemented SAP S/4HANA last year, covering four of their largest sectors, namely Dipped Products, Fentons, Hayleys Agriculture, and Hayleys Advantis. It was such a success that they agreed to do a case study with us.
What they found was when it comes to some functions, they can reduce their dependence on manual activity considerably, and the manpower could be redistributed to better work profiles. It helped them save two days in the monthly closing processes, had a 60% time savings on financial reporting in the Fentons sector, 25% in potential savings from consolidating back office operations, faster shipping turnaround time leading to greater customer satisfaction scores, a clear and accurate view of business performance at subsidiary, sector, and group levels, and a boost in operational efficiency by introducing leading practices in areas such as financial control, manufacturing operations, project controlling, and customer service.
What is your focus on AI in Sri Lanka and what is the potential we have for AI?
AI goes hand in hand with data, because without data, AI is meaningless. The current challenge many organisations have is not the AI part, it’s understanding the data. For this, they have to have data scientists or hire them because AI will do basically what you design it to.
For example, if you call a bank call centre to get some service and at the end, the agent asks you if there is anything else. Instead, how about if the agent says: “Sir, we have this offer for you, would you like to avail it?” Now, that offer would be an offer that is being suggested based on certain learnings where AI can be used – we call it the next best action.
This could be done by profiling you when the call comes, and it can happen instantaneously while the call is ongoing.
It checks what you are doing with the bank and what your demographic and background are. Then, it looks at 100 other people with a similar background and checks what those people have done. It will then use an algorithm and say this is the next best thing for you, as well as the second best thing, and so on. So when the agent offers you something, they are not just randomly doing so and very often, when they hit the nail on the head, you will be keen to take it.
These are things many people here are not doing yet but are trying to do. They haven’t got there yet, but are getting there slowly. We are currently conceptualising this. The trouble we may have is that before doing this, we need all the data.
This data is in multiple systems and we need to consolidate it – we call this master data management (MDM). We create a unique copy first and then we start to correlate all of this including you with similar people. That is where all these algorithms come in. Then, we have a profile of you.
How much of a challenge is it to get Sri Lanka to invest in AI?
When it comes to Sri Lanka, we don’t have the volumes that India or even Singapore has. Despite being a smaller country, Singapore’s volumes are huge as most of the organisations there have expanded regionally.
If you take our volumes here, we are like 1-2% of that. It is very difficult for stakeholders of an organisation to show return on investment because IT deployment costs, whether you do it in Singapore or here, don’t change.
That initial investment has to go in because we are fixing the same problem. That’s where the challenge is. The more globalised organisations are able to do that, which is where the fintechs and all come into play. They develop these models based on their research and become the partners for these organisations to help them develop.
Even initially, when mobile apps were launched internationally, not many people had them in Sri Lanka. It took a while for these to come in as it didn’t have the ability, and then when the mobile apps came in, it was not connected to the backend, so it was not end to end. Only now do we have end-to-end mobile apps.
For these things to work, organisations need to have this agile layout which many in Sri Lanka do not. So for them to build on top of it, it becomes a little more difficult when compared to organisations in Singapore or Malaysia, because they already have those platforms.
Before bringing in AI to Sri Lanka, we have to address more fundamental issues like workplace automation, infrastructure management, security, and end-to-end automation.
What plans do you have for the Sri Lankan market in the coming year?
So we are investing a lot in educating organisations, on security, digitalisation, automation, integration, and basic analytics on setting up data lakes and so on, before we go to AI.
Since the beginning of this year, we have held around 20 events for customers and our partners. So what we are continuously doing is bringing down all the global work we have done, the use cases and experiences, to the people here because this is the time everybody is planning for the next year.
We are in the process of working with the people, and it’s a two-way road. We have to understand what their priorities are and we need to share with them about the things they can do, and develop our plans accordingly.
We see security, analytics, and cloud as major areas, especially with our global acquisition of Red Hat. We are bringing that platform for cloud application development. End-to-end automation and robotic process automation are also going to be big because that’s going to give the productivity component which will directly impact your bottom line.
From a hardware standpoint, the latest systems we are bringing include an on-premises cloud. This is like a hybrid, so that you can even turn your IT function to a profit centre by being able to meter your hardware usage by division.
So you will know who is using it.
The latest thing we are working on is AI in a box called “Power for AI”, which means when you buy that server, it comes with all the tools in-built for you to develop your AI models. We are trying to get one of these to the universities. We are working with the stakeholders to see if we can bring it down at least for the purpose of research.
We are also growing our partner ecosystem and enabling more and more partners to become IBM’s feet on the ground in Sri Lanka.
Photo Pradeep Dambarage