LLM Agent Engineer
Core
Building and refining large language models to act as proactive, multi-step agents that coordinate and optimize complex architectural workflows.
Role type
LLM Agent Engineer
Builds
Production-ready software with clear modular structure for parsing, logic handling, and reporting
Domain
AI automation and complex systems integration
Deliverable
production ML models
Required skills
Python, JavaScript, Go, Java, SQL, multi-turn system interactions, live non-mocked environments, multi-stage coordination workflows, live tool integration (Supabase, Gmail, APIs), persistent state and session tracking, privacy leak detection, authority escalation handling, indirect prompt injection mitigation
Preferred skills
Experience creating multi-stage coordination workflows, connecting agents with live tools, implementing persistent state and session tracking using MEMORY.md, spotting subtle issues like privacy leaks and prompt injection
Technologies
Java, JavaScript, Python, SQL
Responsibilities
Train AI agents by providing human feedback to improve performance and reliability, build and refine large language models to act as proactive multi-step agents, contribute to projects teaching systems to design and optimize complex architectural workflows, review and evaluate complex system behavior, support development of advanced generative systems through analysis and iterative improvement, work on orchestration-focused tasks involving data gathering and structured output, integrate live tools and APIs to solve real-world problems, identify and flag failure modes including privacy risks and prompt injection attempts