Tree is an AI data asset management platform that addresses a gap distinct from agent memory tools and RAG-based knowledge bases. Where those systems focus on retrieval and context injection at runtime, Tree targets the upstream problem: bringing structure to the datasets, model inputs, prompt libraries, and generated outputs that teams accumulate across AI projects. It gives teams a single place to catalog what data they have, track provenance, and ensure the right assets are accessible and reusable across workflows. Built for teams running AI-driven projects who need more organization than a shared drive but less overhead than a full MLOps platform.