Careers at Dukai Studio
Build alongside Dukai.
Dukai Studio places embedded engineers inside growing companies to lay the data foundations and ship the agents that change how those companies operate.
Selected clients
A model cannot run a business it has never seen.
The data layer comes first.
We map how the company really operates and turn that into a single layer, the one source every system and AI tool draws on.
Agents, built on it.
Next we ship the agents that do the actual work, each one scored against a baseline agreed with the client up front.
Model‑agnostic.
We take no referral fees from any lab or vendor. What we build runs on any model and belongs to the client.
01 — Who we look for
Who thrives here.
Data is your thing.
Tables and the messy rest: schemas and spreadsheets, and just as much the PDFs, email chains, and call recordings. You want every bit of it modeled.
You move between stacks.
Any model, any agent framework, any SDK, plus whichever ERP or CRM the client has lived in for years.
You have shipped for real.
Something in production that people rely on daily, and you can explain every choice behind it.
Operations interest you.
Mapping a process and modeling its data is the part you enjoy, not the admin before the real work.
Ambiguity suits you.
A loose brief and an ugly dataset add up to a day well spent.
You want it live.
Whether you sit across from the client or support the engineer who does.
02 — Open positions
No open roles right now.
Introductions open
Every role here does the same work: build the data foundation inside a growing company, then ship the agents that run on it. If that is what you want to be doing when a role opens, send an introduction: a few lines about you and one link to something real you shipped.
03 — What comes with it
What actually makes a difference.
Ownership from the first week.
A client relationship is yours, along with the systems behind it. Not one lane of a roadmap someone else drew up.
Range you will not find elsewhere.
Software, healthcare, construction, eCommerce, manufacturing. A fresh industry and a fresh data problem every few months.
The founders sit with you.
Discovery, scoping, and selling: you pick them up by doing them beside the people who started the studio.
Go where the client is.
Remote by default, with clients spread across the country. Expect stretches on site when the build calls for it.
Our own tooling and agents.
An internal knowledge layer, CRM, and deployment kit, built by the team and sharpened every week. You start from a working stack, not an empty repo.
A real AI budget.
Access to frontier models, the tools you ask for, and tokens to spare. We do not skimp on the thing the work runs on.
Introductions
Introduce yourself.
There are no open roles right now, but introductions are welcome. A few lines about you and one link to something real you shipped is enough; we open that link before anything else.