Remote Prompt Architect for LLM Dialog Systems

Moonfire; New York (NY); On-site; USD 100,000 - 150,000; Full time

Benefits offered by this job

Stock options

Company-paid benefits

Job summary

Moonfire is a fully remote opportunity preferred to be based in the Eastern or Central time zones. You will design prompts and context strategies for large language models, build routing to subagents, and author complex prompts for live conversations. You will document work, teach others, and lead initiatives as you grow with the company. Strong experience with OpenAI/Anthropic, and conversational LLM outputs is valued.

Qualifications

• 3-5 years of experience in AI, LLMs, or related fields.
• Experience prompting and writing for conversational LLM outputs is highly valued.
• Ability to document work and teach others within the company.

Responsibilities

• Design architecture for specialized subagents within live customer conversations.
• Build routing and delegation systems for handling responses, subagents, tools, and human escalation.
• Define contracts between primary agent and subagents: inputs, outputs, confidence, latency, and failure handling.
• Master the dialogue platform and train others on prompting.
• Train Superhumans using prompting, knowledge-base creation, and fine-tuning.
• Prototype development and ai-assisted coding to test prompting workflows.
• Develop multi-media prompting including images and video for LLMs.

Skills

LLM prompting

API integration

Conversation design

Prototyping

Copywriting

OpenAI/Anthropic experience

Job description

Moonfire is a fully remote opportunity preferred to be based in the Eastern or Central time zones. You will design prompts and context strategies for large language models, build routing to subagents, and author complex prompts for live conversations. You will document work, teach others, and lead initiatives as you grow with the company. Strong experience with OpenAI/Anthropic, and conversational LLM outputs is valued.