USW · Research survey
Scientific workflows & AI agents
A survey of working scientists on the problems you solve, the tools you run, how your workflows really unfold — and what an AI agent would have to do to be useful to you.
1/10·Welcome & consent
0%We are researchers building University of Scientific Workflow (USW), a public benchmark that evaluates whether AI agents can carry out real scientific research workflows. To ground the benchmark in reality, we are surveying working scientists about:
- the problems you work on,
- the tools and simulations you use,
- how you design and revise your research workflows, and
- where — if anywhere — AI agents could meaningfully participate.
About 15 minutes+ one optional 5-min module
9 short partsthe last two are optional
Aggregate reportingidentity used only for eligibility
Consent & data use
- Participation is voluntary; you may skip optional questions or stop at any time.
- Your name, email, and ORCID are collected only to verify eligibility and — if you opt in — to contact you. They are stored separately from analysis data and never published. Results are reported in aggregate only.
- This is a pilot research preview run by the USW team (see the Team page). Institutional review information will be posted here before broad distribution.
Your draft is auto-saved in this browser. Questions about the study? See the Team page.