NCS Plans 300 Embedded AI Engineers in Singapore

NCS has opened an AI Mission Centre in Singapore and says it plans to grow its team of AI engineers embedded inside client organisations from about 100 to more than 300 within the next two years.

The Singtel-owned technology services firm announced the centre on 29 October in a press release and a LinkedIn post. Josephine Teo, Singapore’s Minister for Digital Development and Information, officiated at the launch. CNA reported that the centre is in Ang Mo Kio.

The engineers are what NCS calls Intelligentisation Forward-Deployed Engineers, or iFDEs. The company describes them as “domain-expert, vendor-agnostic AI specialists embedded directly within client teams”. The target of more than 300 is a commitment for the next two years, to be met through what NCS calls “reskilled internal talent and new hires”. It has not said how many of the extra 200 or so will be new recruits.

“By growing our current team of 100 iFDEs to more than 300, we are investing in our greatest asset – our people,” chief executive Sam Liew said in the release. Speaking to CNA, Liew said the engineers “will go deep into the customers’ domain and understand the problem, develop solutions around AI, and actually revamp all the core processes within the business”. He told CNA that legacy systems and inconsistent data are among the reasons some companies struggle to adopt AI.

What the centre will do

NCS says the centre will help public and private organisations move from AI experiments to full deployment. Teams drawn from industry specialists, iFDEs and its Sunshine.AI platforms will help clients pick problems, redesign workflows and test AI “securely in live operations”. Successful projects are meant to become reusable products.

The company also announced training plans: AI fluency training for all staff, more than 2,000 AI practitioners being upskilled, and bond-free work-study programmes for university, polytechnic and ITE graduates.

Three memorandums of understanding were signed at the launch:

Changi Airport Group: to enhance and scale iFerret, a runway surveillance system the two developed together that watches for foreign object debris. NCS chief AI officer Edward Chen told CNA it can detect debris as small as 4cm on a runway of about 4km, and that it is already used in the Middle East and Hong Kong.
Institute of Mental Health (IMH): to co-develop digital tools for patient care. The first named project is Bio-ID, which NCS calls “an AI-powered facial recognition solution to strengthen patient identification and support timely awareness of safety considerations in inpatient care settings”. IMH will validate it in its wards.
Life3 Biotech: to co-develop AI tools that optimise the production of micro-organisms.

At the launch, Teo said any country can produce an AI strategy, adding, according to CNA: “ChatGPT can do it for you.” The harder part, she said, is “a real bias for action”.

Facial recognition on psychiatric wards needs clear safeguards

This is TSN’s analysis. The hiring target is the headline, but the IMH project is the part with the most direct effect on people. According to CNA, Chen said the Bio-ID trial has so far involved 180 patients across four wards, and that the agreement would scale it to all 50 IMH wards and to psychiatric wards in other hospitals. Those figures come from Chen via CNA; NCS’s own release and LinkedIn post give no patient numbers.

Correctly identifying patients matters in any hospital. On a psychiatric ward, where some patients may be confused, distressed or unable to confirm who they are, a reliable check could help staff give the right medication and spot safety risks sooner. That is the case NCS makes.

But facial recognition of psychiatric inpatients is sensitive. Patients on these wards may have limited ability to give or refuse consent, and a face template is biometric data that cannot be changed if it leaks. NCS’s announcement says nothing about how images are stored, who can access them, how long they are kept, how patients or families are told, or whether anyone can opt out. It also gives no accuracy figures or results from the four-ward trial. TSN could not find a statement from IMH setting these out by the time of writing. Expanding to 50 wards and other hospitals should come with published answers to those questions.

The jobs plan should be read in the same way. Like New York’s planned AI Jobs Institute and the 4,053 jobs Malaysia expects from approved data centres, the 300 figure is a target. The embedded-engineer model itself is now common: HCLTech and Salesforce have both built AI delivery around forward-deployed engineers. The test for NCS is whether the headcount, and the IMH results, are reported back.