About the work
We are building high-fidelity simulated work environments used to evaluate and improve AI agents on real marketing work. Each workplace reproduces a marketing org's actual tool surface — email, storage, CRM, project management, social media management, web analytics, AEO/SEO, ads, CMS, product analytics and support — populated with realistic documents, dashboards, personas and deliberately planted problems.
You will help design and pressure-test the Social Media environments: the briefs, the artifacts, the judgment calls a strong practitioner would make, and the errors a weaker one would miss.
What you will do
Design task briefs drawn from work you have actually done in social media, including what the task is really testing — the planted issue and the decision a strong practitioner should reach
Specify the documents, dashboards, personas and tool states a realistic version of that task requires
Write or review the reference answer and the criteria that separate a strong response from a plausible-but-wrong one
Review AI agent attempts and judge them against your own standard
Tasks span the capabilities we measure: diagnosing what happened from messy or conflicting data,



prioritizing and making tradeoffs, planning and executing, QA and reconciliation, triage and escalation, research and evaluation, reporting, and orchestrating multi-step work across several tools.
Social Media scope
Social inbox management, campaign and publishing execution, share-of-voice analysis, and the response judgment calls that come with an owned brand channel.
Who we are looking for
Hands-on practitioner experience running brand social — you have owned the account and the inbox, not only managed people who do it
Fluent in a social management platform such as Sprout Social, and comfortable reading native platform analytics
Prior experience building AI training environments, RL environments or simulated case studies is strongly preferred
Rubric Academy or Rubric Bootcamp certification is strongly preferred
Clear written reasoning — much of the value is in explaining why a decision is right
Screening
Applicants complete a short multiple-choice knowledge screener specific to this sub-domain before review.
#J-*****-Ljbffr

📌 Social Media Expert - Ai Environments (New South Wales)
🏢 Mercor
📍 New South Wales