AI Engineer – LLM System Prompt Development AI Trainer, $50–$60/hour

Project Overview: Join a growing community of professionals advancing the next wave of AI. As an AI Trainer, you’ll play a hands-on role by analyzing and providing feedback on data to improve LLM performance, helping ensure that the next generation of AI technology is accurate and trustworthy. We are seeking a skilled AI Engineer to work as a project consultant in our AI Labor Marketplace. This is not a full-time employment position — you will be engaged as an expert project consultant on a contract basis. Location: U.S.-based experts only Engagement: Part-time, project-based expert evaluation work Work Type: Remote Project Summary: Contributors will design, build, and validate an LLM system prompt capable of generating realistic, schema-compliant Microsoft Power Platform tenant state JSONs from scenario descriptions. The work focuses on prompt engineering, model testing, and output validation, ensuring the system prompt reliably produces high-quality synthetic tenant data across different LLM models for use in AI agent evaluation. Consultant Engagement Terms: This is a project-based consultant role. Consultants will be paid on a per-project basis; hourly rates are estimates based on anticipated completion time. Consultants control their own schedule, provide their own tools, and may simultaneously provide services to other vendors/employers (subject to those vendors’ allowances). Responsibilities: Contributors will: • Design and iterate on an LLM system prompt capable of generating detailed tenant state JSONs from a given scenario document • Build and configure a test framework to execute the system prompt using Opus 4.5 and/or ChatGPT 5.2 • Develop model-specific prompt variants to account for differences in LLM behavior and output quality • Validate generated tenant states for schema compliance, referential integrity, and internal consistency • Ensure each generated tenant state accurately reflects the conditions described in its source scenario • Track and document gaps between scenario requirements and what the current tenant state schema can represent • Package and deliver all system prompts, LLM configuration files, generated tenant states, and gap documentation Expected Outcomes: • A finalized, tested system prompt with model-specific variants for Opus 4.5 and ChatGPT 5.2 capable of reliably generating schema-compliant tenant states from scenario documents • A working test framework that can be built, run, and iterated on • One tenant state JSON per scenario, conforming to the schema and representing a realistic tenant • A documented log of any schema gaps identified during generation Qualifications: • Strong hands-on experience with LLM prompt engineering for structured output generation • Demonstrated ability to generate valid, complex JSON reliably from LLMs • Experience debugging and validating schema-based data structures • Comfortable working with deeply nested, interrelated data models • Familiarity with software frameworks for running and iterating on LLM-based generation pipelines • U.S.-based and able to work independently in a remote, asynchronous setting Nice to have: • Familiarity with Microsoft Power Platform concepts (environments, apps, flows, licensing, DLP policies, governance structures) • Experience working with enterprise tenant configurations or cloud administration data models  LinkedIn Global Data Privacy Notice for Job Candidates: https://legal.linkedin.com/content/dam/legal/Global-Data-Privacy-Notice_LI_TalentCommunity.pdf