As organisations increasingly embrace the hybrid multicloud environment, the demand for a robust yet agile infrastructure is increasing. This is one of the reasons why ST Engineering entered into a Memorandum of Understanding with Nutanix in 2023, whereby the two companies plan to collaborate on creating cutting-edge hybrid cloud technologies, among others.
However, to support these innovations, there should be enough skilled manpower at the helm. With the global talent crunch in technology roles, ST Engineering and Nutanix had to focus on another important aspect of the business — training and development.
Raja Gopal, ST Engineering’s General Manager for Mission Software & Services, sat down with Frontier Enterprise on the sidelines of the Nutanix .NEXT 2024 conference in Barcelona. Gopal discussed the company’s strategy for solving the global tech talent shortage, their AI thrust, and how the enterprise world will look like in the next five to 10 years, especially with regard to cloud deployment.
Could you share a bit of background about your MoU with Nutanix?
There are two trends behind that move. The first one was about three or four years back, when there was a major rush for customers to adopt the public cloud environment. Thereafter there were two reasons for the ensuing pushback— one was the rising costs, and second was data security. Over time, we are starting to see most of our customers adopting a hybrid cloud strategy, where you keep your data sovereign in a private environment, while utilising the services of a public cloud and having the best of both worlds. And this is where the partnership grew between ST Engineering and Nutanix.
The second trend is the AI trend. If you look at the Gartner Report from maybe three years ago, there was no mention of generative AI, as a sort of big, powerful initiative. Of course, there are some smaller efforts on trying to do machine learning and data analytics, but last year, generative AI spread everywhere. So when we looked at this from a hybrid cloud perspective, we started to see a shift toward a hybrid multi cloud environment.
We started to build a private cloud with Nutanix in 2019, which is meant for our customers and ourselves. We progressed very fast with this project, and last year, when we signed the MOU, it was really to build two things in two different areas: the first one is in the hybrid multi cloud environment, where you have disparate management platforms causing nightmares for customers. This meant we were required to build a common cloud management platform, and to be able to manage end to end systems across these clouds through a single pane.
The second one is the workforce. Currently, we have about 500 cloud engineers, 300 of which are Nutanix-trained. This is the largest cloud engineering workforce in Singapore and Southeast Asia. Nutanix helped us curate a curriculum that is beyond academic and is more skills-based. We are seeing a major push among many of our engineers to get the various certifications in Nutanix. There’s only one Nutanix expert in the region, out of the total four globally, and we have that one expert at ST Engineering. There are multiple Nutanix masters that we have at ST Engineering, so we are building a very strong practice together with the engineering support from Nutanix.
The reality is that we don’t have enough Singaporean ICT workforce, this forced us to explore alternate resources. We then set up an academy called Digital Academy to train not only the cloud engineers, but also the non-ICT workforce, such as the ones from broadcasting, marketing, and economics to take up cloud engineering. In fact, when they join us for the training, we train them up for about nine months before we send them back into the workforce.
We’ve also developed a skills-based rewards programme rather than an academic one, so that we can train non-degree holders and if these resources are good after three years, then we will remunerate them in the same way we remunerate a degree holder.
Moreover, we have mothers who want to go back to the workforce after career breaks, and people who have been in the industry for some time and now want to be retrained. With all these initiatives, we are able to support the major demand that we are seeing from our customers in this hybrid multi cloud movement.
How does your collaboration with Nutanix go about?
In the case of the private cloud, we actually designed, built, and maintained it, from an infrastructure point of view. We have built the entire cloud environment from scratch, because while there is the Nutanix platform, you must also build the services and integrate it with third party services. The whole design phase in fact comes in before you start building.
The maintenance part is interesting because we are building automation. You can see a huge reduction of Day 2 operation resources, from 60 down to 16. When you’re building this infrastructure, you can press one button and you’re going to execute many of the hardening and patchworks required at the Day 2 level. However, the one that I think we are trying to really pursue is using AI to run Day 2 operations. This means taking away the human element as much as possible in many instances where the machine is able to see and execute certain actions. Actually we are involved in this entire end to end process.
Now, there is one more step which is migrating the workloads. At ST Engineering, we have the cloud team as well as the application team. We have a team that is now taking the applications and is either changing the whole software and pushing it up, or plainly just migrating it. For migration, we are doing both lift and shift and refactoring.
Our Digital Academy not only builds Nutanix experts, but also trains people in Kubernetes, OpenShift, or Splunk, to name a few. Those are the other expertise that we are building in order to make sure that we have an ecosystem of a workforce that is capable of supporting this end-to-end solutioning and delivery for our customers.
How do you collaborate with GovTech to create a skilled workforce?
GovTech has this program called continuous learning. What GovTech has proposed, and what we are doing at this moment involves three months deep training, and six months, on-the-job training for learners. So when an individual who has learned software programming, for example, comes to the academy, they learn something else, like agile development. Then we put them into on-the-job training, where they run together with the project teams, and they understand the different roles. For example, what can a UI-UX specialist or scrum master do under the scheme? They can learn software development, for example, under the programme. By the time they graduate from the academy, and they go into the workforce these employees are very confident that they know what they’re supposed to do.
Are you currently focusing on AI skills development over at ST Engineering?
The AI wave has hit all industries. If you look at the history of ST Engineering, we service a broad range of customers and industries like public safety and security, government agencies, aviation, healthcare, maritime, and transportation. Now, this whole set of domain areas each have unique requirements of how AI should be applied in these environments. Therefore, ST Engineering is building a strong AI core. Last year, we did a major reorganisation whereby we created the Group Engineerings Centre as well as a Group Technology Centre. The idea behind these centres was to build these cores with AI features inside them. They centralise and build a very strong AI practice to be able to support all our customers in these domains.
We are currently molding two kinds of people: one is people who understand how to manage and manipulate data, and hence, data engineering is a key requirement. The second one is people capable of building the learning models, and putting it into applications, so that it can be used for a better purpose.
We are actually also looking at overseas resources to augment the Singapore workforce, especially those that are not so sensitive in nature. I believe the global resources out there will help us speed up the rate at which we can build up this workforce.
In terms of the AI learning models, are you working on your own algorithms?
We use the ones available out there, but not exclusively one reason being the nature of natural language. Singapore is actually very unique. You have Singlish, Malay, Chinese, and then a mix of everything. When the language models taken from open source are applied in a production environment, many of the trials showed that they may not necessarily work. We have built our own LLM, together with the National University of Singapore and it is being used in our government agency environments. And today, it works very well, because you are really able to do code switching. This is because in one moment someone is speaking in English and then in the next moment, he switches to Singlish and from there he goes to Malay, finally returning back to English. This code switching has to be done very fast for the customer to use it in their production environment.
What do you think is the ideal or final state of the cloud environment?
I look at it from a time horizon perspective. In two to five years, I think we will still be in the building stage. Organisations will be investing and building up this separate infrastructure. Five years and beyond, I think two things are going to happen: one is that we would have built up enough capability for seamless shifting. Second, I think quantum will kick in.
You can divide quantum based on two criteria— the first one is selective quantum for application workloads, and the second is quantum for cybersecurity, such as encryption. This entire idea of quantum key distribution in order to protect infrastructure is one key area that we think is going to kick in quite fast. The second one is quantum in application development to be able to match the speed of things. Our assessment is that quantum is pretty far away, but it’s actually starting to speed up.
