About this role
Design and implement prompt architectures, guardrails, evaluation frameworks, and RAG systems to ground LLM-powered character interactions; integrate multi-provider LLM APIs and collaborate with creative and brand teams to ensure consistent, safe, and measurable character fidelity.
Skills for this role
Prompt engineeringSystem prompt designFew-shot techniquesChain-of-thought reasoningMulti-turn conversation managementStructured output designRetrieval-Augmented Generation (RAG)Embedding modelsVector storesVector retrievalChunking strategiesRetrieval evaluationOpenAI APIGoogle GeminiAnthropicLLM evaluation frameworksAdversarial testingSecurity testingPrompt versioningRegression testingModel evaluationAgentic designPythonAWSText-to-Speech (TTS)Speech-to-Text (STT)Guardrails frameworksFine-tuningReinforcement Learning from Human Feedback (RLHF)Cross-functional collaboration