AI Engineer

Summary

Join our team as a skilled software engineer specializing in Large Language Models (LLMs). You will be responsible for designing, planning, and implementing software solutions, collaborating with other engineers, and advocating for product quality and performance. Your expertise in Python, FastAPI, Langchain, and observability will be crucial for building and maintaining our LLM stack. You will also contribute to prompt engineering, LLM behavior tuning, and ensuring efficient LLM call management.

Requirements

• Advanced Python, experienced with FastAPI
• Deep knowledge of Langchain
• Observability and debugging experience (Langfuse preferred)
• Strong database knowledge (SQL, Redis)
• Experience with asynchronous architectures
• Amazon Web Services knowledge
• SCRUM Methodologies practitioner
• Prompt engineering and LLM behavior tuning

Responsibilities

• Participate in the processes of designing and planning software solutions
• Suggest ideas, new solutions, or improvements to the current technology systems
• Collabowith the Pro and other stakeholders within Engineering (Frontend, UX, etc.)
• Document the technical solutions with diagrams and the necessary documents for easy understanding by the other technical areas of the company
• Advocate for improvements to product quality, security, and performance
• Craft code that meets our internal standards for style, maintainability, and best practices for a high-scale web environment
• Maintain and advocate for these standards through code review
• Recognize impediments to our efficiency as a team ("technical debt"), propose and implemize LLM stack

Preferred Qualifications

• Experience with Node.js and NestJS
• Understanding of agent proxy architectures
• Familiarity with Websockets and streaming LLM responses
• Knowledge of vector search and RAG techniques (e.g., Pinecone, Qdrant)
• Familiarity with LLM evaluation methods and tools (e.g., LangSmith, OpenAI evals)
• Awareness of LLM-related security and compliance concerns
• Ability to manage cost, token usage, and latency of LLM calls
• Product thinking around user experience with LLM-powered features
• Experience with fine-tuning or hosting open-source models