01 · When nobody could be in the room
This work belongs to a very specific moment: COVID. Showrooms were shut, career fairs were cancelled, shoe stores were empty. Every client arrived with a version of the same problem — our customers can't experience us in person anymore.
At Accenture Song (then Entropia), the answer we kept building was presence without place: virtual worlds and augmented reality that ran in an ordinary browser. The hard constraint, and the thing that made this genuinely difficult, was that none of it could assume a headset. No Meta Quest, no wearables — immersion had to happen on a flat screen, or it didn't happen at all.
02 · Four brands, one playbook
For Accenture itself, we built Careerverse — a virtual career fair where candidates walked into a digital version of our office and talked to recruiters as if they'd visited in person. Built on React and Babylon.js over microservices, it lifted candidate participation by more than 20% — the cancelled career fair, outperformed.
For BMW Malaysia, the problem was test drives. We recreated the full Genting Highlands road — the climb Malaysian drivers actually dream about — as a virtual driving experience in Three.js, so a customer at home could feel what that car would do on that mountain.
For Nike, sales were down and stores were closed, so we put the fitting room in the customer's camera: a WebAR experience on 8thWall where you point your phone at your own feet and see the shoes on your own legs — digital twins of the real pairs.
Alongside those: an interactive eBrochure for Telekom Malaysia (React, AWS, Google Analytics), and a launch campaign for Netflix's Money Heist. Plus the emerging-tech edge of the same era — ETH-based NFTs, e-wallet integration, gamified experiences and digital-twin planning.
03 · My role: the person in the middle
Immersive delivery is genuinely multi-disciplinary: a graphics team building models and environments, a frontend team making them run at interactive frame rates in a browser, a backend team wiring services underneath. Someone has to hold those three to one vision — and hold that vision to what the client actually meant.
That was my seat. I was the customer-facing technical lead: I took the brief from the brand, turned it into something each team could build, and kept the threads aligned until launch. And I'll say it plainly — 3D was new territory for me. I was learning the domain while leading its delivery, which is its own kind of leadership test: you can't bluff your way through a frame-rate problem in front of a graphics team.
04 · The hard part: immersion on a flat screen
A headset hands you immersion for free; a browser tab gives you nothing. Making someone feel a mountain drive, or trust how a shoe looks on their own foot, through a flat screen — that was the real engineering and design problem under every one of these projects.
It disciplined every decision: model budgets and lighting that read as real at browser frame rates, camera work that sells motion without inducing it, interactions that feel physical with nothing but a mouse or a phone. The lesson that stuck with me: the technology was never the point — the feeling was. The tech only had to be good enough to disappear.
05 · What it shows
I won't oversell this one — it isn't a metrics story, and I'm not going to dress it up as one. It's a range and leadership story: marquee global brands, a brand-new technical domain, three specialist teams to coordinate, hard launch dates attached to marketing campaigns, and a genuinely awkward constraint.
What it proves is the shape of leader I am when the ground is unfamiliar: I can stand between a demanding client and specialist teams, learn fast enough to make real technical calls, and ship. The domain was new; the way of working is the same one I bring everywhere.