James Chia’s introduction to workplace learning arrived in three files. A young civil-service trainee, he was handed the documents and told to read them. There was little explanation to help him understand what he was reading, or why. An organisation had supplied information. A newcomer was left to supply the meaning.
It is an unglamorous beginning for a technology business: no garage, no midnight revelation, just some paperwork doing a rather poor impression of a teacher. Yet the memory stayed with Chia. Years later, when he co-founded ArcLab with Steven Chan, he would work on a problem those files had made familiar: how to help someone learn what they need, when they can actually use it.
Today, that question takes him well beyond an office desk. ArcLab serves employers training people on shop floors, in restaurants, in logistics and on construction sites. These are workers whose jobs happen somewhere specific, with customers waiting or tasks to finish. Their working day does not conveniently pause for a lengthy presentation.
The teacher who refused to stay on the syllabus
Chia had already experienced the other kind of education. In Primary 5, his form teacher, Mr Liang, taught Chinese and mathematics. The combination impressed him: a teacher educated in Chinese was explaining mathematics in English. The adult at the front of the room had done some difficult learning himself.
Mr Liang also wandered far beyond the immediate business of getting children through examinations. He brought ancient China, scientific ideas, Tang poetry and Leonardo da Vinci into the classroom. The subject was nominally on the timetable; the world kept slipping in through the door.
Chia remembers the intellectual curiosity more clearly than the academic material. That distinction matters to his later work. Getting information in front of a learner is relatively easy. Giving that learner a reason to care takes a different kind of effort. His tribute to Mr Liang returns to the idea of educators helping people make connections and learn how to learn.
“Our role is to bring this knowledge out, and put it in a very simple, effective delivery format.”
James Chia, on working with employers
A banker’s rather expensive missing autograph
Before ArcLab, Chia worked at the Monetary Authority of Singapore. His career included reserve management in Singapore and London, putting him inside the financial world during the global financial crisis. He is also a Cornell alumnus. These were substantial institutions with well-established ways of doing things.
One London assignment placed him at a dinner with Lee Kuan Yew, Singapore’s then-high commissioner to Britain and Bank of England governor Mervyn King. After the official business, Chia raised manufacturing as a way for financial centres such as Singapore and London to diversify their economies.
He had met Lee before, receiving the Prime Minister’s Book Prize as a schoolboy. The books remained on his shelves. The second encounter left a smaller, more domestic regret: he had wanted an autograph for his wife, but felt asking would be improper while attending as a member of MAS staff.
There is something nicely revealing about that missed request. Chia later contrasted his rule-following public-service self with the more flexible founder he became. Even in a room full of economic policy, the detail that survives is a husband deciding he really ought not to ask.
Leaving the dealing room
The transition began with family. Chia left MAS to become a stay-at-home father, with no next job arranged. He remembers changing nappies and taking long bus journeys with his elder child. Time with his daughters became the thing he had chosen to buy with the income he gave up.
He later marked a curious symmetry: eight years and eight months at MAS, followed by eight years and eight months outside it. The neatness belonged to the calendar. The business journey was considerably less tidy.
His first venture taught him that being an accomplished employee offered no automatic qualification for entrepreneurship. He compared running a technology business without technical knowledge to opening a chicken-rice stall without knowing how to cook. It is a useful image, particularly in Singapore, where the customer is unlikely to be impressed by a persuasive slide about lunch.
The experience changed how he approached a second attempt. He put more effort into understanding customers and learning practical skills. Intelligence and good intentions would need some help from execution.
First, a game. Then, a smaller lesson.
That first education business, Innervative, created Oikonopolis, a city-building-style game teaching teenagers economics and personal finance. Chia dates the game’s creation to 2013. It gave him experience of designing a learning product, along with the less enjoyable experience of making business mistakes.
When he described the beginnings of ArcLab in 2018, he set out three requirements for learning: it should engage people, use their time efficiently and help them understand through involvement. Those requirements came out of a product journey, rather than a sudden enthusiasm for putting a classroom on a screen.
A game can involve a learner in making choices. A workplace tool has another constraint: the learner may simply need to perform one task well. ArcLab’s direction took Chia toward smaller units of instruction that could fit around a working day. The design challenge was to preserve useful engagement while removing unnecessary demands on someone’s time.
The founder becomes the student
An early prototype also provided Chia with a lesson in who gets to be the expert. Bing, a young developer he later celebrated in an account of ArcLab’s beginnings, took the team’s functional ideas and designs and built a proof of concept in six weeks.
Chia admired Bing’s ability to break complicated requirements into manageable pieces. The prototype gave the team something companies could try. Feedback could now concern an actual tool, rather than an idea everyone might politely agree sounded promising.
Chia began learning JavaScript and Python through Codecademy. His aim was to understand code and work better with developers; Bing helped when he got stuck. The founder writing about lifelong learning had placed himself in the learner’s chair, with a younger colleague available to explain the difficult bits.
That story helps explain his interest in routes into work beyond conventional academic filters. He saw someone build something useful. The result made its own argument. In 2019, Bing returned to work with ArcLab’s development team after completing National Service.
ArcLab incorporated; an early product followed.
Full platform launched in July, welcoming paying customers.
EduSpaze’s first cohort, HolonIQ recognition and Bisk Ventures investment.
The old phone earns its keep
One of Chia’s more instructive product decisions lived in his pocket. In 2021, he described using a five-year-old, 16GB iPhone 6S. It gave him a practical way to check whether ArcLab’s learning modules would work on the modest devices some workers owned.
A product could look perfectly agreeable on an expensive new phone while making life harder for its intended audience. His old handset kept that possibility close at hand. A technology company’s founder had found a business use for declining an upgrade.
He also named the Rocky films as an inspiration: the underdog who trains, persists and cares about his family. His Tottenham allegiance fits his own account of rooting for the less fashionable contender. He has followed the club since 1994, a commitment with ample opportunity to practise perseverance.
These details make the accessibility argument tangible. A worker’s phone is part of the design brief. So are the worker’s language, available time and familiarity with technology. Chia’s admiration for underdogs comes with some quite ordinary engineering consequences.
Everyone takes a turn on the chat
By 2024, Chia was explaining another habit at ArcLab: team members, including the CEO, took turns answering customer questions online. The company was serving larger organisations, but he wanted the people building its software to hear directly from the people trying to use it.
He described learning business through mistakes, without coming from a business family. His conclusion was disarmingly plain: “to be a good person first, before we can be a good business person”. It is a statement about conduct, with the customer conversation as one place to practise it.
In a later interview that year, he described visiting customers’ workplaces to watch staff use the platform. He also discussed translation tools and the importance of having native speakers review machine-generated language. The work involved learning where people encountered friction, then improving something the whole customer base could use.
His ambition is to make workforce training easier to reach across languages and countries. ArcLab’s presence in Singapore and Malaysia gives that ambition a regional footing, while its focus remains on people whose work offers little resemblance to a software developer’s day.
The person serving the bubble tea
Chia’s interest in frontline work also extends to the hours around training. In an April 2024 commentary, he asked readers considering flexible work to think about the person preparing their bubble tea. Working from home offers that employee few practical possibilities. Flexible hours or fewer shifts might.
He acknowledged the operational complications: rosters still need filling, handovers cost money and larger teams need more training. His argument asked employers to consider the potential benefits alongside those costs, including access to people who could work only particular hours.
It is the kind of problem his career has taught him to notice. A policy or a training programme can appear complete from the desk where it was written. Its usefulness becomes clearer at the counter, on the floor, or in the hands of someone reading it for the first time.
Three files once left a young officer to work out the context for himself. Chia now spends his working life asking how much of that burden a better-designed lesson can remove.