Rapid Prototyping AI Architect (Live Server/Brownfield)

We are looking for a brownfield LLM First Agentic Full Stack Software Developer and Architect. We are rapidly prototyping an MVP and looking for someone experienced with LLM first development in fully established code bases. This is a sprint, not a slow and steady focus on sustainable clean code (yet). We are using OCI Instanced servers in a currently live environment so care must be maintained to preserve uptime in pushing development. PLEASE READ: We are NOT building an AI chatbot, a RAG pipeline, or an AI voice assistant. The software is a traditional web CRM (PHP/EspoCRM). "LLM First" means we are looking for a developer who uses AI IDEs (Antigravity, Cursor, Claude Code, Aider) as their primary tool to write traditional server architecture at 10x speed. We are looking for developers who are specialized in guiding AI Coding Agents through the development pipeline at a professional level. Keeping control of the machine as it writes near 100% of the syntax without allowing it to fall out of control pushing changes without permission, creating scripts that break other features or make changes that weren’t permitted to it, or re writing features multiple times over. You are an experienced LLM pilot, the LLM writes the syntax.