Contribute human expertise to AI training, evaluation or data work. Check the official source for project requirements.
About the role
Project Overview
We are sourcing independent Finance and Insurance Specialists to provide their expertise for an AI benchmark evaluation project. As AI models increasingly generate professional-grade financial and insurance deliverables, their accuracy relies entirely on robust, expert-crafted training data. The objective of this project is to autonomously produce high-quality evaluation tasks, strong prompts, and clear, well-structured rubrics that generate clean, reliable data for model training.
Project Deliverables & Scope
Operate autonomously to design complex evaluation frameworks and provide structured training data. Expected deliverables include:
Task & Prompt Creation: Generating realistic, high-quality prompts that compel the AI model to produce complex, professional-grade deliverables specific to the finance and insurance sectors.
Rubric Development: Writing clear, well-structured evaluation rubrics with criteria that are highly specific, non-ambiguous, and easy to score.
Benchmark Evaluation Data Generation: Producing clean, reliable training data that directly aids in the evaluation and refinement of AI models handling industry-specific tasks.
Quality Assurance & Fact-Checking: Ensuring all generated tasks and scoring criteria reflect strict, real-world industry standards and regulatory realities.
Required Expertise
To successfully fulfill the deliverables of this project, Contractors must possess deep industry knowledge to craft realistic professional scenarios.
Core skillset includes:
Demonstrable professional expertise within the Finance and Insurance sectors, with a deep understanding of industry standards, terminology, and high-level deliverables.
Strong writing and prompt generation skills, with the ability to design highly realistic, complex task scenarios for AI evaluation. …
Official source: job-boards.eu.greenhouse.io. Details can change; the platform's page is the final word. How we check listings.
