How NCS is empowering Singapore’s public sector with AI

“For the public sector, the Singapore government has been very progressive in terms of adopting technology. The AI use cases that they are looking at can also be very interesting,” says Wynthia Goh, Senior Partner, Global Co-lead, NEXT at NCS.

Governments in the Asia Pacific region continue to focus on digitizing public sector services as they look to serve the community better. In Singapore, digitalization in the public sector is one of the faster growing sectors in the region.

With AI now seemingly the next step in perfecting government services, developing and deploying GenAI use cases that can improve the public sector is seemingly an imperative for the country. The Singaporean government has already committed to invest up to SG$28 billion from 2021 to 2025 to drive digitalization in the island state, which AI and quantum computing solutions included as well.

More recently, in May 2024, the Singaporean government said it would spend about S$3.3 billion on ICT procurement, with S$2.1 billion focused on modernizing infrastructure to meet evolving challenges and maintain public trust.

Enabling all this to happen will be Singapore’s tech ecosystem. While major tech vendors have already invested, partnered and committed to the government to help with the development of AI in the country, local partners and distributors play an important role in connecting the dots and bridging the gap in the public sector.

One example is the strategic partnership between NCS and AWS. The Singaporean technology partner collaborated with AWS on multiple impactful public sector projects such as the recent pivotal work on Singapore Ministry of Manpower (MoM)’s GenAI-powered Contact Centre. NCS has also recently been awarded AWS’ APJ Public Sector Consulting Partner of the Year – a testament to the strength of the ongoing partnership.

In a conversation with CRN Asia, Wynthia Goh, Senior Partner, Global Co-lead, NEXT at NCS discussed the company’s work in Singapore in delivering GenAI capabilities for the public sector.

What are the challenges companies are facing in their AI journey?

There have been some changes in the last couple of years. There's an evolution and it reflects the overall maturity of the organization. Take the example of Gen AI. A year and a half ago, there's a flurry of experimentation. Many companies are rushing to test out using Gen AI, developing various types of POCs.

Two years down the road today, we see a more nuanced and differentiated approach depending on the company's maturity. Certainly, there are still companies early in the journey and working on some POCs, but we have also seen our clients, who had done that previous round of innovation, moving on to look at more complicated use cases that has the potential of also having a more impactful outcome.

And of course, there are clients that are a lot more mature in various aspects and set up with their organization. They are already at a point where they are deploying AI applications. The conversations with them now are on how to build up a continuous loop of innovation and how to continue to keep tap of the new innovation or new emerging technology that's coming through in that space as well as how that informs the further execution of their AI strategy.

How can organizations deal with the shortage of skill sets to work on AI and other emerging technologies?

The shortage of AI talent is quite a global trend and there are a few ways that companies can look at it. To properly adopt AI, an organization needs to make sure that a good cross-section of their people is involved in the journey and they may be involved in various ways depending on what are the stages and what are the activities that the organization wants to execute.

The first thing that we typically would recommend is the organization needs to enable their own people to at minimum understand what the capabilities and potential of AI are, learn to be able to use the tools and the organization needs to enable their people with the right AI tools in order to be able to do their job better.

Now beyond that internal enablement, the organization also needs to look at how they would be adopting AI, how they would be adopting AI applications into their business, their operations and how that may transform the way they operate or how that may develop new opportunities to be able to pursue new areas of growth.

In those areas, what our clients should look at and I usually advise them is to look at a combination of how to build internal capabilities and how to have access to partners or to build an ecosystem to have access to the type of talent that they need, bearing in mind that AI itself can be quite a broad area.

When we talk about AI talent, from one project to the next, the nature of AI talent you need may also be different. This is also where when you work with a partner like AWS, they also offer training on their capabilities. The relationship we have with AWS is quite a holistic and comprehensive one.

Back in July 2024, we had announced with AWS the launch of NCS x AWS GenAI Center of Excellence for Public Good for the APAC region. If you look at the partnership, we not only jointly work with clients on potential AI use cases, but on both ends for NCS and AWS. We are also enabling sufficient AI expertise in knowing the latest of AWS technology. This puts us in a good position to have the capabilities to help our clients explore the most innovative use cases.

How different is it to work with the public sector on their AI journey as compared to the other sectors and industries?

Every organization and every industry client have their different considerations. For the public sector, the Singapore government has been very progressive in terms of adopting technology. If anything, working for the public sector is something that's a lot more impactful because we are working with clients that themselves have gone through quite a fair bit of a digital journey. The AI use cases that they are looking at can also be very interesting.

We are very happy that AWS awarded us their consulting partner for the region award. It's the first time that a Singapore company has won this award. And it really reflects three things. First, the business impact that we delivered in our relationship with AWS. Secondly, the capabilities that we built, and we keep building, to be in our best position to be able to support our clients. And thirdly, which we are proud of is the innovation that we've delivered. And one big area of that innovation has been the work with the public sector.

For example, the work on Singapore’s Ministry of Manpower’s contact center. The contact center, at a national level, receives all of the incoming calls and queries related to manpower employment issues. As you can imagine, given that broad topic that this call center is covering, call center agents may be receiving calls that in order to address the queries that's coming through, they actually internally need to check across multiple sources of information.

The job and the queries that they deal with can be quite complex. At the same time, the work that they have been doing requires quite a fair bit of searching through multiple data sources to be able to find the right composite response to deal with a customer query.

So, we did a couple of things with them.

We explored with them how can GenAI be deployed to support the contact center agents in their role. And there were a few things that we did. First, we put in place a GenAI search that allows the agent to more easily search through those multiple data sources and be able to generate a potential response given the query that the agent is dealing with on the live call. This assists the agent to help them be able to do their job better and faster as well as elevate part of the work that they are doing now.

The second part is we also put in place a hyperlocal speech-to-text engine that NCS had created from working on a training data set of a large number of Singapore call center call data. With this hyperlocal speech-to-text engine, we are able to achieve transcription accuracy that's above 90%, much higher than out-of-the-box solutions that are typically available to most clients. With this, we are able to do a few things.

First, we help the call center agent to be able to get a more accurate and faster transcription of what happened. And then that also allows for the post-call analysis, which then leads to the third part, which is how we elevate the call center agent's post-call work. Because after every call, there is certain paperwork and documentation that they need to do.

With that highly accurate transcription, we are now able to generate those post-call summaries in under 10 seconds. And then what we and Minister of Manpower discovered is that we are first able to reduce those non-call works by more than 50%. Also, because generative search is able to help the agents search through and identify potential responses to the queries, we are then also able to reduce the call duration for each call by 12%.

Prior to deployment, we went through a process where we did a POC with a smaller number of agents to test out the flow of the process and also to, very important, get feedback from the call center agents because this application is designed to help them and to elevate their work.

From that POC, we eventually scaled out to all agents for the contact center. When we design these GenAI applications, it's quite important to say what are the benefits that we are trying to deliver, which customer, including internal customer, we are trying to serve. How does AI help solve what problem? And in this case, it's very clear to us that we have designed this to elevate certain pain points of work for the call center agent.

What we ended up achieving is not only from a productivity perspective. We see these productivity gains, but really also happier call center agents. And of course, hopefully happier customers as well.

The work that we have done for the Ministry of Manpower Contact Centre, from the learning and best practice through the AWS network, will also be shared to the rest of the world. Our hyperlocal speech-to-text engine is also an NCS asset that we have built out there.

If you think of the challenges that the hyperlocal speech-to-text engine is trying to solve, you can see that that problem and that challenge may exist in other parts of the world. And through the partnership with AWS, this is also a way where innovation that's coming through here in NCS can also be leveraged by clients elsewhere in the world as well.

When working with the public sector, how do you ensure that you're keeping to data privacy and regulatory requirements when developing all these use cases?

I think data is a very important foundation for all of the work that we do and all of the work that the clients want to do, especially in the area of AI.

NCS has a long history of working with the public sector and large enterprises. We're very familiar with the kind of data security compliance standards that must be adhered to, and different industries may have different regulations, and so different rules and compliance requirements may apply.

Hence, being fully aware and knowledgeable about that and designing it with data security in mind is very important. In fact, when we announced our AI-Digital Resilience (AI+DR) Matrix a few months back, under digital resilience, we put data governance as a very important element that companies need to get right.

In fact, data governance is one of the foundations. When an organization is clear about their data governance, their data classification, which data can be trusted and how data will and will not be used, then there is confidence upstream on the AI application that is built, and then there's confidence for end users, citizens, consumers, to use the AI applications.