With every project we ship, we try to become a better version of the studio we were last quarter. We keep ourselves abreast of the tools, technologies and trends that move through this industry — and we put time and capital behind the ones we believe in, well before a client asks us to.
Most of our R&D output starts here: arrows, boxes, a question someone has been turning over for a week. Once a quarter the whole studio steps out of client work for five days, and the diagram on the wall is the only deliverable that matters.
Read the essays →Most studios in Hong Kong are paid to deliver last year’s stack on this year’s deadlines. There is little internal pressure to learn anything new before a client asks for it, and the work quietly drifts a year or two behind.
We try to invert that. A fixed share of every working week, and one full week per quarter, is reserved for work the studio pays for itself — internal tools, AI experiments, infrastructure rewrites, essays, conference talks. Each engineer’s ten-percent learning time runs in parallel.
The output ends up in client engagements the next quarter — cheaper, faster, and with fewer surprises than if we had learned on someone else’s clock.
A working area we keep returning to: small, opinionated tooling for measuring whether a model or prompt change actually improves the output for the specific task at hand — because off-the-shelf eval frameworks tend to be either too academic or too tied to one vendor.
Reusable building blocks for the patterns that recur across HK retail, F&B, and finance apps: bilingual content management, MTR-friendly nav, HKID validation, Octopus / FPS integrations, member-tier UX. Re-used quietly across engagements so each new project doesn't start from zero.
An ongoing investigation: how far can open-weights models (Qwen, Llama, Mistral) be pushed on dense legal & financial retrieval, running entirely on hardware the client controls? The answer matters for any HK advisory firm working with regulated data.
The internal scaffolding that lets one or two of our engineers sit inside a client's operation and ship faster than a larger off-site team would: provisioning, observability, internal-admin scaffolding, customer-support tooling — all the unglamorous stuff that compounds across engagements.
We don't fund speculative bets on five-year horizons. Every lab project must plausibly show up in a real client engagement within twelve months. If it can't, it isn't a fit for us — that is for venture capital, not for a services studio.
A small win on a boring stack (Postgres, a cron job, a clean queue) usually beats a flashy framework that nobody on the team can debug at 3am. Our internal tools are deliberately under-engineered.
Internal tools, libraries, and templates are written as if a new joiner will inherit them on day one. The discipline matters more than the publishing — well-documented internal artefacts compound across engagements; un-documented ones quietly rot.
An essay shipped is research published. Every senior engineer is expected to publish one piece of long-form writing per year — on what they learned, what they got wrong, or what they think the rest of HK should be doing.
The small, opinionated tooling from Focus 01. Used internally on every AI engagement to keep model and prompt changes honestly measured.
A small TypeScript form & validation kit that ships English and Traditional Chinese error messages in lock-step. The validation rules our HK clients keep re-discovering they need.
A Traditional-Chinese-first tokeniser tuned for HK product search. Outperforms the default Postgres + jieba pipeline on small HK corpora.
Reference flows for HK payment methods — Octopus, FPS, AliPay HK, WeChat Pay HK — that have actually shipped in production apps, with their gotchas documented.
The CLI we use to spin up a new client project: Next.js + Postgres + auth + a sensible analytics setup, ready to ship in an hour.
How we write code, write English, write Chinese, write commits, write postmortems. Maintained as a living internal doc so new joiners and embedded engineers can ramp up on day one rather than picking it up by osmosis.
Open to internships, applied research collaborations, and guest lectures with HK universities and tertiary programmes — on terms that respect everyone's IP. Reach out if there's a fit.
We work across the major HK-region clouds (AWS, GCP, Azure HK / SG regions) and with both proprietary and open-weights model vendors, choosing the combination that fits the client's data-residency and budget constraints.
We support HK SME associations and trade groups with pro-bono engineering hours and occasional speaking on practical AI adoption for F&B, retail, and logistics. Topics where we can be honest, not promotional.
For projects needing local presence beyond Hong Kong — Taipei, Tokyo, Singapore — we collaborate selectively with trusted partner studios in those cities. We refer work both ways when the fit is right.
If you’re a researcher, a vendor, an industry association or another studio looking to collaborate — write to us. We’re selective, but we always read.