Zipporin is seeking a talented and disciplined Prompt Systems Engineer to lead the design and implementation of resilient, production-ready AI interactions. In this 100% remote role based in the United States, you will bridge foundational large language models and enterprise application logic, transforming stochastic model outputs into reliable, structured, and performant system features.
Key Responsibilities:
- Design, version-control, and scale structured prompt architectures, chain-of-thought workflows, and autonomous agent systems.
- Construct rigorous, automated evaluation harnesses to measure LLM performance, hallucination rates, token efficiency, and output consistency.
- Collaborate with backend engineering teams to integrate LLM pipelines with Retrieval-Augmented Generation (RAG) systems and vector search indices.
- Implement guardrails, input/output validation, and red-teaming mechanisms to secure models against prompt injection and malicious exploitation.
- Profile and optimize prompt latency, context window utilization, and inference API costs across diverse foundational and open-source models.
Requirements:
- 3+ years of professional software engineering experience with deep technical proficiency in Python.
- Proven track record building production-grade prompt systems and tooling using frameworks like LangChain, LlamaIndex, or DSPy.
- Demonstrated experience operationalizing LLMs (OpenAI, Anthropic, Bedrock, or open-source Hugging Face models) in enterprise environments.
- Solid understanding of vector databases (e.g., Pinecone, Qdrant, Weaviate) and semantic search retrieval techniques.
- Strong background in testing methodologies, continuous integration, and data-driven evaluation for generative AI.