Senior AI Engineer

What You’ll Do

• Arcitect and build LLM-powered systems — design retrieval workflows, context management, agent prompts, and structured output pipelines.
• Orchestrate AI workflows using tools like LangChain, LlamaIndex, or similar frameworks, integrating them with product APIs and backend services.
• Drive prompt engineering and iteration — refine prompts, templates, and context strategies to meet product quality and reliability goals.
• Manage real-world evaluation metrics — measure usefulness, factual correctness, latency, and UX impact vs. classic accuracy alone.
• Collaborate across functions — work closely with product, platform, and backend teams to ensure seamless integration.
• Develop reliable, scalable deployments — focus on performance, cost efficiency, and observability in production environments.

Who You Are

• Experienced in building real LLM applications — you’ve shipped systems that use large models meaningfully.
• Strong software engineering skills — Python/TypeScript, API design, backend integration, and cloud deployment.
• Tool fluency — comfortable with RAG, vector databases (e.g., Pinecone/Weaviate), workflow frameworks (LangChain, Dust), and related tooling.
• Architectural thinker — you can diagram end-to-end solutions incorporating context windows, caching strategies, tool calls, and multi-step reasoning.
• Product-oriented — you care not just that the AI works, but that it delivers value safely and reliably to users.

Basic Qualifications

• 6+ years of software engineering experience, with at least 1+ year working building AI in production.
• BS Degree in Computer Science or related field. Masters degree is a plus.
• Demonstrated use of LLMs in production workflows or complex prototypes.
• Strong coding ability in Python or equivalent; familiarity with backend frameworks and cloud services.
• Experience with API integrations, database systems, and scalable architectures.
• Experience with multi-modal models or multi-agent system design.
• Familiarity with AI safety guardrails, hallucination mitigation, and structured output enforcement.
• Knowledge of vector DBs, RAG architectures, and prompt lifecycle tooling.