Jakarta, Indonesia
AI development for Jakarta teams, handed over in writing
One shop, Vancouver, British Columbia. Jakarta sits on the other side of a full day, so nothing about this engagement runs on meetings. It runs on written handoffs, logged decisions and review that happens while the other side is asleep.
- base
- vancouver, bc, canada (UTC-08:00)
- market
- Jakarta, Indonesia
- your clock
- UTC+07:00 (Asia/Jakarta)
- overlap
- none on the same date, written handoff instead
- presence
- remote only, no office in Jakarta
No shared working day, so the loop is written
Jakarta is 15 hours ahead of Vancouver. When we start at 09:00 it is 00:00 on the next day for you, and when we stop at 17:00 it is 08:00 on the next day. On the same calendar date there is no hour where both sides are at work.
That is not an apology. A loop that turns once every twenty-four hours, in writing, is a faster way to build software than a calendar full of half-attended calls, and it forces the decisions into a form somebody can read six months later.
- Everything built during our day lands in one written note before we close the laptop. No status meeting stands in for it.
- Your morning starts with that note and a set of pull requests to review, which is a better start than a calendar invitation.
- Every decision worth arguing about is written down where the code is, with the reasoning and the date attached.
- Questions come with a recommendation, because a question that needs a reply before work continues costs a full day here.
One live call a week, and it is scheduled honestly: 18:00 in Vancouver is 09:00 in Jakarta on the next day. That hour is outside our working day and inside yours. We take the awkward end of it, because it is our job to be reachable.
What Jakarta buyers usually bring us
The record for Jakarta lists two industries. Here is the part of each one that turns into an engagement, and the service line it lands under.
consumer technology
Consumer scale means the eval harness matters more than the demo: a one-in-a-thousand bad answer is a support queue, not an anecdote.
Acceptance criteria are fixed before each weekly increment, which turns "is it working" into a question with a yes or a no.
Closest service line: AI products, built end to end
financial services
Research, filings and internal policy are the corpus. Hybrid search with metadata filters as hard constraints, plus span citations, or it does not ship.
Chunking follows the structure of the document instead of a character count, which is the whole difference between a useful index and an expensive one.
Closest service line: Knowledge bases and retrieval
Contracts, currency and where the data sits
Invoices are in US dollars. Contracts are in English, and where the data sits is agreed before ingestion rather than after.
The rest is the same everywhere: we scope it, we build a thin slice against your real data, and we hand back something your own team can maintain. No retainer you cannot cancel and no seat licences.
The five lines we sell
Same five lines everywhere, Jakarta included. Most engagements are one of them, sometimes two.
- AI products, built end to end
- Knowledge bases and retrieval
- Codebase and team optimization
- Training and enablement
- Forward-deployed engineering
The languages, models and infrastructure underneath all five are listed on the Magic Ship stack page, with what each one is good for and where it is the wrong choice.
Other markets in Asia
More of Asia, all of it worked remotely from Vancouver.
Start a project
Write down the problem in whatever form you have it. You get a straight answer about scope, timeline and whether this is work we should be doing at all.