Get a reply from this employer
Your resume and cover letter, tailored to exactly what they’re hiring for.
Armature is seeking a researcher to own public research on how coding agents pick and use tools, and to manage the evals system that underpins those results. The role requires turning raw data into clear findings and shipping those insights into the product with engineers.
You design studies and evals, automate research runs, and write findings that developers want to read. Daily work centers on agent-driven research and moving fast without waiting for instructions in Paris.
The role asks for someone to own both the public research on how coding agents pick and use tools, and the evals system that backs those results. A strong candidate can take raw data to a clear finding alone, designs studies and evals, and ships what works into the product alongside other engineers. The posting stresses daily building through coding agents and moving forward without waiting for instructions.
The work sits directly on how coding agents behave, so AI is the subject rather than a side tool. Much of the research running and analysis can be automated, which lets each study go faster than the last. What stays human is choosing the questions worth asking, judging what the data means, and writing findings that developers want to read.
A person here would gain hands on practice studying how agents discover and select developer tools, and how models and harnesses differ. They would learn to design evals that hold up against real customer questions and to test new eval methods against the company's own data. Building this means running agent driven studies end to end and turning the results into shipped product features.
This is our own read, written from the employer's posting and the company's Spotlight. The employer's posting is the source of truth for the role.
Armature helps developer tools evaluate how autonomous coding agents discover, select, and interact with their products, per the company.