AI adoption starts with people, not technology, says Temus
According to Sutowo Wong, Managing Director of AI x Data at Temus, there is still a lot enablement needed on how companies can train employees to use AI as well as how they can get more value from the core use cases.
For Sutowo Wong, Managing Director of AI x Data at Temus, the Singapore-based digital transformation services consultancy has grown tremendously this year not just in terms of scale but also in their geographical reach.
Speaking to CRN Asia during Tech Week Singapore, Wong shared that the recent launch of their AI Foundry as well as the strategic investment into ThinkingMachines a couple of weeks ago only reinforces their presence in the industry.
With AI Foundry focused on developing core use cases for enterprises, Wong explained that customers in the industry are moving from the periphery where the risk is lower to the core of the organization.
“That's a very good sign that we are now seeing as more enterprises wanting to get real value out of AI. Within the AI Foundry, we are building together a multidisciplinary team all the way from people who can look at AI strategy down to the product managers, the AI engineers, domain experts or specialists. This allows us to be able to work very closely with our clients across the whole transformation agenda taking the domain lens end-to-end from Vision to Value, a framework we have in Temus,” he said.
On the strategic investment into ThinkingMachines, Wong said the company brings more than a decade of experience in applied AI and data, which is very valuable to Temus. What makes it more exciting for Temus is that ThinkingMachines is part of the geographical reach they are hoping to achieve.
Given that ThinkingMachines is also an OpenAI partner in the region, Wong explained that ThinkingMachines focuses a lot on that as they have a whole suite of services all the way from training and enablement.
For Wong, the whole reason for Temus to exist is all around capability development.
“It's not just about bringing people who are ready-made, who can then go and work with the client. It's also about how we develop and train our people. We have a good training curriculum to be able to expose them and train them on various skill sets,” he said.
AI enablement
Interestingly, Wong shared that there is still a lot enablement needed on how companies can train employees to use AI as well as how they can get more value from the core use cases.
“That's why there's a lot of alignment in terms of our approach, in terms of our ability to actually deliver for the client. One trend that we are seeing is going from access to operationalization. Everybody has access to the same models. The key is operationalization. We see a lot of requests from clients coming to us and say that, how do I operationalize? That includes going into the data to understand what kind of data is necessary for AI to be grounded so that then it gives you useful answers instead of generic answers.
Wong added that this comes along with the whole slew of modernization of data platforms, improving data quality, down to how do organizations put the right governance in place.
“For example, do I have agent identity? Can I manage that? Do they follow the access rights and controls that are actually given to the users who have created the agents? Do I have the ability to have a kill switch if I sense that something is not right? Am I able to audit and observe what an AI agent is doing? So, a lot of the work has moved into operationalization,” he added,
Wong also pointed out that customers are also still going back to the basics on the value of AI. There were a lot of the early experimentations around the use cases that are low risk. Wong shared that the low-risk use cases naturally are not the ones that are going to move the needle.
“There's a lot more discussion around what is the core of the organization? How do I use AI to update that? Take for example, in asset management, valuation is a very critical component or workflow for asset management company. So how do we use AI to uplift the productivity around valuation? So, I see a lot of that happening also where clients are moving towards the core processes that actually makes a lot of difference to them,” he added.
At the same time, Wong also believes that there are still a lot of concerns of trust in AI.
“How do you build trust within the workflow, within the infrastructure? How do I get people to build up their confidence? And that has to do with that broad-based enablement that we are doing. This is where ThinkingMachines also come in. Building that confidence is important because while we can have top-down direction to say that I focus on the core, but it's up to the people to actually do the delivery. But if the people actually who are supposed to do the delivery do not have the confidence, that adoption will not be there. And we all know that the lead indicator for ROI is actually adoption,” Wong said.
“We are very cognizant that the reason why we are brought in is to provide that neutral objective view,” Wong said.
While Temus are partners in the AI journey for enterprises both large and small, Wong pointed out that the goal is all about execution.
“When we are looking at the needs of the client, we are very clear that we don't come in with a bias to say that I want to implement a certain style of solution. So it's always starting with the client needs. And based on their needs and requirements, what are the candidate solutions out there? What makes sense for them on a variety of dimensions such as cost and vendor lock-in. How do we actually try to minimize that? Although you can't really avoid it altogether by minimizing the vendor lock-in. What are the things that you will need from an operational standpoint? Because it's important for the enterprises to be able to actually run things on their own as well,” he said.
He added that Temus is very clear about not just building capability within Temus but to work with clients to build capabilities for them too.
“When it comes to the selection of the right technology, it's also about operational sovereignty. Are they able to operationalize and continue to work on them after Temus has left? So a lot of the work with our partners are around getting an understanding of how they can meet the needs of our clients,” he concluded.