1—2
senior engineers, on-site
Forward-Deployed Engineering is the engagement shape we've been refining since 2014 — and it's how Accord works best. One to two senior engineers working closely with your Hong Kong operation on a weekly cadence — remote-first, on-site when it genuinely helps — writing software against the actual workflow rather than the brief you wrote in a vacuum. The fastest path we know from messy ops to production software you and your team can read.
Most "outsourced" engagements end with the vendor halfway across town, behind WhatsApp/Slack, behind a project manager, behind a third invoice you didn't expect. Forward-deployed is the opposite shape.
We work alongside your team on a recurring weekly cadence — remote-first, on-site when it genuinely helps. We learn your workflow by watching it run. We ship small internal tools that compound across months, in the languages your team actually speaks. We don't go "black-box" — every change is visible to your team by end of day.
Forward-Deployed Engineering is AccordHK's flagship engagement shape: one to two senior engineers embedded inside a Hong Kong operation on a weekly cadence, writing software against the actual workflow rather than a brief written in a vacuum. Engagements run 6 weeks minimum, priced per-engineer-month on working days, with handover built in from week one.
Free 35-min intro call, then a short discovery — remote or on-site — where one of our engineers maps how your operation actually runs. We leave you a written scope, a recommended team shape, and a fixed working-day quote. If you decide not to proceed, you keep what we wrote.
The first piece of software ships inside two to three weeks — chosen because it’s the smallest thing that removes a real bottleneck. Boring stack on purpose. Your team uses it on a Monday, we adjust it on a Tuesday.
Same engineers, same standup, same demo slot every week. The internal tools compound — a dispatch view, an admin, a data pipe, a kitchen tablet. Each one paid for the engagement that built it; the next one runs cheaper.
When you’re ready to run things alone, the handover is already built in — runbooks, code documentation, recorded walk-throughs, your in-house developer on every PR review. We stay reachable. We don’t hold the keys.
We don’t publish a sticker price on this page because Custom / FDE engagements require us to understand your scope first. Every quote we produce is built up the same way:
(working days × engineer day rate) + capped expenses. The day rate is fixed for the engagement. Scope changes are agreed in writing before any extra work happens — no surprise invoices, ever.
The first 2 weeks are explicitly scope-only: we shadow your operation, leave a written scope and recommended team shape, and only then quote the full engagement. If you don’t proceed past scope-only, you keep the scope doc.
Shorter than that, FDE doesn’t produce real leverage — it’s just expensive consulting.
We bill in working days, not hours — no hourly-creep, no clock-watching.
Travel, infra, third-party tooling — all itemised, all capped per month, all approved before charge.
If scope shifts mid-engagement, we write it down, agree it together, then do it — never the other order.
Typically 3 to 6 small, focused internal tools by month four — each one removing a real bottleneck in your operation.
A written ops manual for every tool: how to run it, how to recover from common failures, who to call.
A short written summary after each working cycle — what shipped, what changed, what’s next — so your team always knows where things stand.
For AI work, an eval harness so changes can be measured. For ops work, structured logs + a sensible dashboard.
From week one, we agree what handover looks like: which engineer becomes your in-house pair, what the rollover criteria are, when we step out.
When the engagement ends, the people you worked with stay reachable on email and Slack. Not a hotline, not a ticket queue — the same humans.
It isn't staff augmentation (we're not filling a hiring gap) and it isn't pure consulting (we ship code, not slide decks). One to two senior engineers work alongside your team on a weekly cadence, writing software against your actual workflow — and if we're no longer useful, we hand over and leave.
6 weeks. Shorter than that, forward-deployed engineering doesn't produce real leverage — it becomes expensive consulting instead. Most engagements run in cadence for several months, with handover built in from week one.
Per-engineer-month, working-day calculated: (working days × engineer day rate) + capped expenses, fixed for the engagement. The first 2 weeks are scope-only — we shadow your operation, leave a written scope, and only then quote the full engagement.
When you need 10+ engineers immediately, when there's no operational floor for us to embed into (pure greenfield product work), when you need 24/7 on-call support, or when the brief is already fully specified — in those cases a fixed-price Project Build or a larger studio is the better fit, and we'll tell you so.
Tell us what you're trying to fix. The first call is free, 35 minutes, no deck. We'll tell you honestly whether FDE is the right shape — or point you to a better partner if it isn't.