DatoricCareers
We are not actively
hiring right now.
Datoric is a small team. We grow by intent rather than by headcount target, and we open roles only when there is a specific problem that needs a specific person. There are no open opportunities at this time.
Status
No open roles
Last updated April 2026.
We post openings on this page when they exist. We do not post aspirational job descriptions.
When we do hire
Roles open up around specific research questions, not around fixed job ladders.
When we identify a question that the current team is not the right shape to answer, we describe the question publicly and invite people whose work suggests they would be the right shape to answer it. The role exists because the question exists, not the other way around.
If that approach sounds familiar, you may be the kind of person we want to know about before there is a role to attach to.
How we work
Four principles that shape who we work with.
№ 01
Empirical first
We measure what we claim. Every research note we publish is reproducible, with bootstrap confidence intervals on every number and the dataset attached.
№ 02
Small teams, full ownership
We do not staff layers between researchers and the work. The person designing the benchmark writes the paper, ships the dataset, and answers the email when something breaks.
№ 03
Boring tools, sharp questions
We use the simplest stack that lets us run the experiment. Our edge is the question we are asking, not the framework we are using to ask it.
№ 04
Long horizon
We care about whether a finding is still true in five years. That bias shapes everything: what we publish, what we measure, and who we want to work with.
People we want to know
Even with no role open, we keep a short list of people we are tracking.
- 01Researchers who care about licensed, ethically sourced data as a first-class object of study
- 02Engineers who can build evaluation harnesses that survive contact with real models
- 03Linguists, clinicians, lawyers, and other domain experts who want to translate professional reasoning into training signal
- 04ML practitioners who have opinions about why models break in production
Introduce yourself
If you would want to work
here when there is a role,
tell us now.
Send a short note about what you are working on and what you would want to work on. We read everything. We respond when there is something to respond to.
research@datoric.comWe acknowledge every message within two weeks.