Malaysia Can Turn Construction Expertise Into Home-grown AI Products -- Archiboost Engineer
By Harizah Hanim Mohamed
KUALA LUMPUR, Nov 5 (Bernama) -- Malaysia’s artificial intelligence (AI) investment boom has strengthened its digital foundation, with the next opportunity lying in bringing industry firms and local developers together to turn Malaysian expertise into useful AI products that can compete globally.
Malaysian software engineer Yi Sien (Ian) Ku sees the construction sector as an opportunity for Malaysia to develop its own AI products, drawing on his experience developing AI software for the architecture, engineering and construction (AEC) industry in Silicon Valley.
Ku is the founding engineer and first technical hire at San Francisco-based Archiboost AI, which has raised US$6.2 million in funding from US investors.
“I would like to see more Malaysian companies using AI throughout their operations, alongside Malaysian developers building the products that make that possible,” he told Bernama.
At a recent media briefing, World Bank lead economist for Malaysia Apurva Sanghi said increased demand for AI-related products drove more than 70 per cent of Malaysia’s overall export growth in early 2026.
Sanghi said growth outside those products remained weak, raising questions about Malaysia’s exposure to a slowdown in AI demand and how much AI-driven growth was benefiting Malaysians.
The World Bank also identified property developers, construction companies and utility providers among those capturing the most immediate economic gains from AI-driven growth.
Noting Malaysia’s strong foundation in digitalisation and engineering expertise, Ku cited Gamuda Bhd’s autonomous tunnelling technology, developed by local engineers, which uses AI to control key operations of tunnel boring machines.
However, Ku sees further opportunities for AI earlier in a project, at the design stage, before construction begins.
These include helping professionals find project information, review designs and identify problems before they become costly to resolve on site.
FINDING PROBLEMS BEFORE CONSTRUCTION
Design coordination provides one example. Architects and engineering teams often work in parallel, with decisions in one discipline affecting others.
“A drawing can appear reasonable on its own while conflicting with another team’s work,” Ku said.
At Archiboost AI, Ku played an integral role in designing and building an AI review system that connects information across drawings, specifications and project requirements to identify potential design conflicts.
In one project, he said the system flagged an underground electrical trench that appeared to run through part of the building’s concrete foundation. It identified the conflict by comparing the trench and foundation locations and depths with the relevant structural notes.
“If a conflict like this reaches construction, it could mean rerouting the trench, revising the design or interrupting work while the team agrees on a solution.”
“Identifying it in the plans gives the engineers an opportunity to resolve it earlier, before it becomes a problem on site,” he added.
Once construction has taken place, a correction may also require the removal of completed work, replacement materials and additional labour.
Ku said one client estimated that correcting a single issue flagged by Archiboost could have cost hundreds of thousands of US dollars had it been discovered only during inspection.
“On a large project, costs like these can add up to millions of dollars,” he said.
Ku said those consequences also shape what professionals need from AI. Before changing a design, architects and engineers must understand why the system has raised a concern and examine the supporting evidence.
To support that review, Ku and the Archiboost team developed the system so the AI can mark the precise areas of concern on the drawings, explain its findings and recommend corrections.
He said a tool must also fit the team’s daily work, including connecting with existing software and project information and providing checks suited to the project stage.
BUILDING PRODUCTS WITH INDUSTRY PARTNERS
Ku said Malaysian industry firms can help local software developers understand professional needs, build useful products and test them in practice.
“The starting point is to identify where staff lose time or struggle to make decisions. Industry professionals can explain the task, the information it requires and the consequences of getting it wrong.
“Developers can then assess how often the problem occurs and whether solving it could support a business,” he said.
Helping staff find and understand project information offers one opportunity beyond design review.
A team may need to locate a material specification, determine whether a requirement applies to part of a building or check an earlier decision. Answering these questions can require reading drawings alongside specifications and building codes.
AI can help connect those sources and explain the relevant requirements, while directing professionals to evidence they can check. Firms could also take a more data-driven approach by using records and lessons from previous projects to guide decisions on current work.
Ku said Malaysian architects, engineers and contractors can contribute the knowledge of local requirements and working practices needed to develop such tools.
“Malaysian teams need industry partners who will test ideas with them and give feedback grounded in the work,” he said.
Experienced practitioners can assess the results, while decision-makers can commit people and time to trials.
These trials give firms a basis for deciding whether to adopt a tool while helping developers establish whether it improves the work enough for customers to pay for it.
Turning that demand into a sustainable business also requires developers to understand delivery costs, including computing, integrations, maintenance and customer support, Ku said.
Ku also believes Malaysia can build more successful AI companies by making it easier for talented people to find the support needed to start and grow a business.
“Coming from Malaysia, I was struck by how easily a conversation in San Francisco could lead to a co-founder introduction, an investor meeting or a customer willing to test something new,” he said.
He said Malaysia could strengthen those pathways by connecting software teams with experienced founders, early-stage investors and companies willing to pilot local products.
TAKING MALAYSIAN-BUILT AI PRODUCTS TO GLOBAL MARKETS
As those businesses grow, Ku said they need to understand customers in the markets they intend to serve.
“Working at Archiboost in the US gives me direct exposure to how firms evaluate new products and make buying decisions,” he said.
Ku said working with customers across markets helps him distinguish common industry problems from local needs and build relationships with people who understand those markets.
For Malaysian developers, professionals working abroad could provide market knowledge and introductions, while Malaysian firms with international operations could connect local teams with potential customers and partners.
For construction software, this means understanding which parts of a product can serve the same task across markets and which must adapt to local requirements or the software a customer uses.
“For regional growth, I would look for a common problem across markets and a product that can adapt to their differences,” he said.
“I would measure progress by the products developed here, the skills our teams acquire and the Malaysian companies that grow alongside the data centres,” he said.
-- BERNAMA