About Adzzat
Adzzat is an applied AI research lab. We turn enterprise data and human expertise into training data and evaluations for frontier AI labs.
The role
Own research and production of coding-focused benchmarks built around real AI failure cases.
Location: Mumbai, India
Non-negotiable
You have created a benchmark, or done real hands-on work on one. Sole ownership is not required, but your own contribution must be explicit, with an artifact or public work we can inspect. This matters more to us than your degree or your title.
What you'll do
Build multiple benchmarks, with a repeatable task-design, validation and review process that keeps quality high.
Run our current benchmarking projects with AI labs.
Study agent runs, group the failures and turn them into objectively graded tasks.
Define what makes an evaluation practical: correctness,
objective grading and difficulty.
Guard against answer leakage and gaming, and measure genuine improvement.
Guide the researchers who write and review tasks.
Solid skills
Coding-agent evals such as Terminal-Bench and SWE-bench style tasks
Experimental judgment: what a benchmark does and does not measure
Reproducible pipelines
Quality control across many outputs
Ideal profile
PhD student or PhD in AI, machine learning or a related field. This is a preference, not a filter.
How to apply
Apply with your resume and a link to the benchmark, repo or paper you worked on. Add one line on what you built.
adzzatlabs.com
📌 Head Of Research, Benchmarks Mumbai (India)
🏢 Adzzat
📍 India