Prompt Engineer (Remote - W2 Only)

Prompt Engineer (Remote - W2 Only)

We are seeking an excellent and highly skilled Prompt Engineer to join our team. This is a remote, W2-only position designed for technical experts who understand the nuances of Large Language Models (LLMs) and can bridge the gap between human intent and machine execution.

The ideal candidate isn't just someone who "talks" to AI; you are a master of structured inputs, chain-of-thought reasoning, and iterative optimization. You will be responsible for building, testing, and refining prompts that power high-performance applications.

Key Responsibilities

• Prompt Architecture: Design, develop, and refine high-quality prompts using advanced techniques (Few-shot, Chain-of-Thought, ReAct, etc.) to ensure accurate and reliable LLM outputs.

• Performance Optimization: Evaluate and benchmark prompt performance across different models (e.g., GPT-4, Claude, Gemini, Llama) to minimize hallucinations and maximize utility.

• Workflow Integration: Collaborate with software engineers to integrate prompts into production environments and automated workflows.

• Content Governance: Establish guardrails and safety protocols within prompts to ensure ethical AI behavior and brand alignment.

• Continuous Research: Stay at the forefront of the rapidly evolving AI landscape, implementing the latest prompt engineering methodologies and toolsets.

Qualifications & Requirements

• W2 Status: This role is strictly for W2 candidates. We are not accepting C2C or 1099 arrangements at this time.

• Technical Excellence: Deep understanding of how LLMs function, including their strengths, limitations, and the impact of temperature, Top-P, and other hyperparameters.

• Linguistic Precision: Exceptional command of the English language with the ability to write clear, concise, and logically structured instructions.

• Analytical Mindset: Strong data-driven approach to testing prompts and analyzing results to drive iterative improvements.

• Bonus Skills: Experience with Python, API integrations, or vector databases (Pinecone, Milvus, etc.) is highly preferred.