MLflow

MLflow is an open-source platform developed by Databricks to manage the end-to-end machine learning lifecycle. Provides unified solution for traditional ML, deep learning, LLMs, agents, and generative AI. Components include Tracking (experiments with parameters/results), Models (deploy across platforms), Model Registry (staging to production with versioning), and AI agent evaluation/tracing. MLflow 3.x expanded focus on generative AI and LLM workloads. Over 30 million monthly downloads, 850+ contributors, powering ML workflows for thousands of enterprises. Positioned as foundational MLOps layer in 2025 industry roadmaps.