Public Sector AI Training & Evaluation Specialist

Risedatalabs; Tulsa, Northern (OK, KY); Hybrid; USD 60,000 - 90,000; Full time

Job summary

Rise Data Labs is seeking a Government & Public Sector AI Training Specialist to craft realistic policy tasks and rubrics and to grade AI outputs against them. You’ll work on government-focused benchmarks and contribute to improving AI reliability in public programs.

You’ll author policy briefs, regulatory reviews, and grant or budget documentation, using redacted sources and formal documents as references, while calibrating scoring with experts.

Qualifications

• Prior hands-on experience in public administration, policy analysis, regulatory affairs, or government program management.
• Comfort reviewing detailed formal/regulatory documents.
• Prior experience with AI model training, evaluation, or benchmark development is a plus.

Responsibilities

• Create realistic work scenarios based on your own professional experience — e.g., policy briefs, regulatory compliance reviews, public program reports, grant or budget documentation
• Supply the kinds of reference materials you'd normally work from (redacted reports, policy documents, program data)
• Write detailed rubrics specifying exactly what a correct, professional deliverable requires — accurate analysis, proper regulatory language, correct formatting
• Grade AI-generated deliverables (often policy documents or reports) against your rubric
• Join calibration sessions with other experts to ensure consistent scoring across the benchmark

Skills

Public administration

Policy analysis

Regulatory affairs

Government program management

AI model training

Evaluation

Benchmark development

Job description

Rise Data Labs is seeking a Government & Public Sector AI Training Specialist to craft realistic policy tasks and rubrics and to grade AI outputs against them. You’ll work on government-focused benchmarks and contribute to improving AI reliability in public programs.

You’ll author policy briefs, regulatory reviews, and grant or budget documentation, using redacted sources and formal documents as references, while calibrating scoring with experts.