FlowiseAI/Flowise

Drag-and-drop tooling for LLM workflows has gotten crowded, but Flowise has earned its place through sheer breadth and stability. It lets you assemble conversational agents, RAG pipelines, and multi-step tool-calling flows by connecting nodes on a visual canvas — no orchestration boilerplate required. Under the hood it leans on LangChain and LlamaIndex primitives, so the abstractions map to real concepts rather than hiding them entirely. You get built-in support for dozens of LLM providers, vector stores, memory modules, document loaders, and API endpoints out of the box. The self-hosted deployment is straightforward enough that a solo founder can have something running in an afternoon. The visual interface also makes it genuinely useful for showing stakeholders or clients what an agent is actually doing step by step. The honest reservation: complex branching logic and custom business rules can get messy fast on the canvas, and you will eventually hit walls that push you back toward code anyway. -> Best for: technical PMs or early-stage founders prototyping AI agent features before committing to a custom backend.