Job Title: LLM Engineer / Prompt & Agent Specialist Location: Remote Type: Full-time Role Purpose We are looking for a hands-on LLM Engineer to design, iterate, and evaluate prompts and agent behaviors. You will own system prompts, tool/function-call specs, schemas, guardrails, and evaluation processes to optimize accuracy, safety, latency, and cost for production LLM workflows. Key Responsibilities Build and optimize prompts for structured outputs and tool/function calling. Design and implement evaluation frameworks to measure performance, errors, hallucinations, latency, and cost. Develop safety mechanisms, guardrails, and adversarial/red-team testing for LLM workflows. Define clear human/machine handoffs with deterministic flows. Collaborate with engineering and product teams to improve LLM-based systems in production. Requirements 2+ years building with LLMs in production (or 4–6+ years engineering with LLM ownership). Expert in prompt engineering for structured outputs and tool/function calling. Fluency with LLM APIs (OpenAI, Anthropic, etc.). Strong evaluation skills: golden sets, rubric/auto-judge scoring, error taxonomies. Experience with safety, moderation, and rule-based constraints in LLMs. Nice to Have Experience with RAG, embeddings, and vector stores (Upstash, Pinecone). Fluency in Typescript. Familiarity with agent frameworks (Mastra, Langgraph, etc.). Experience with eval/prompt tooling and version management (OpenAI Evals, DeepEval). Knowledge of A/B testing, fine-tuning, and human-in-the-loop feedback loops. Observability and monitoring experience for LLM systems.