AI Tinkerers KL — co-directing the chapter that became KL’s AI practitioner pipeline.

What KL’s AI community needed

The AI Tinkerers community exists to give practitioners a place to share what they’re actually building — not slide decks about AI strategy, but working code, real demos, real failures, real lessons. When we kicked off the Kuala Lumpur chapter, the city had AI conversation in plenty of places: corporate panels, vendor-led summits, consulting roundtables. What it didn’t have was a regular room where the people shipping AI products could find each other and learn from peers, with no agenda beyond the work itself.

That gap was the thing worth filling.

The kickoff, the cadence, and the hackathons

We kicked the chapter off in August 2024. The first events were monthly — one room, one shape, mixed audience. That worked for getting to a critical mass of people, but as the community grew it became clear that the builders and the business operators in the room wanted different things from a meetup. The builders wanted technical depth: someone walking through their RAG architecture, the fine-tuning run that worked, the eval setup they actually shipped. The operators wanted product judgement: what AI features customers paid for, where adoption stalled, how to cost a model in production.

Rather than try to serve both in the same room, we shifted to a twice-monthly cadence with two distinct tracks. One event a month for builders, focused on technical work. One event a month for the business side, focused on product and go-to-market. The split doubled the value the community delivered without doubling the cost of running it — the people who wanted both came to both, the people who only wanted one got the one they wanted, and neither side felt they were sitting through content meant for the other room.

Alongside the regular events, we ran practitioner hackathons — focused, time-boxed builds where the community shipped real things together rather than just talking about them. The hackathons pulled in the builders who don’t always show up for evening meetups, and gave the chapter a different kind of energy from the panel-and-demo format.

The partner layer

A community without partners is a meetup. A community with the right partners becomes infrastructure. We built that layer deliberately, across a few overlapping fronts.

The capital and founder layer. Antler’s KL presence was a regular co-conspirator — founders we hosted often turned into Antler-funded startups, and Antler-funded startups regularly came through our rooms. The community and the venture studio fed each other.

The infrastructure and applied-AI layer. YTL AI Labs anchored the conversation about AI compute at industrial scale, including some of the largest AI data-centre development in the region. Grab brought applied-AI engineering depth from one of the strongest AI teams in Southeast Asia. TDCX brought the deployment-at-scale lens — what AI actually looks like running across multiple markets at customer-experience volume.

The financial-services layer. Paynet, the operator of Malaysia’s national payments infrastructure, anchored conversations about deploying AI inside regulated financial rails. Moneylion brought practitioner depth on consumer-facing AI fintech products with regional reach.

The academic layer. Asia Business School anchored the business-track conversation with the rigour of a school setting alongside practitioner depth.

Plus a wider list of partners and supporters who showed up across events, hackathons, and sponsor sessions through the year. Each partnership came with a real reason. The community got access. The partners got direct connection to practitioners they couldn’t reach through marketing. The currency was relevance, not sponsorship.

What it became

The community started returning value beyond what its founders put in.

AI Tinkerers KL became a direct pipeline into consulting work, client introductions, and funded startup collaborations. AI workshop opportunities at a Big 4 firm and major Malaysian enterprises — including Paynet — came partly out of these relationships. The OtonocoAI engagement intersected with the same network. Founders found co-founders, engineers found their next role, and people who’d been working alone discovered they had peers. That’s the right outcome for a practitioner community — the room creates flow, and the flow turns into work.

The honest part

AI Tinkerers KL was a volunteer role. The work was real — kickoff, two-track cadence, hackathons, partnerships, sponsor conversations, talent flow — and unpaid through the year. The team worked on monetisation in parallel, because community work that depends on indefinite goodwill eventually breaks. Building a sustainable model alongside running the chapter was its own quiet project, less visible than the events themselves but central to whether the format could last past the energy of its first year.

That’s the reality of practitioner community in a city that doesn’t yet have venture-funded models for the format. We named it openly to ourselves and to partners. That honesty was part of why the partnerships held.