Flowise

About Flowise

Flowise was created in 2023 as an open-source, Apache 2.0-licensed tool for building LLM-powered applications through a visual drag-and-drop interface. It wraps LangChain, LlamaIndex, and other AI libraries behind a no-code/low-code interface, enabling teams to compose RAG (retrieval-augmented generation) pipelines, conversational agents, and multi-step AI workflows without writing code. Users create 'chatflows' by connecting nodes representing LLMs (from any provider—OpenAI, Anthropic, Groq, local Ollama models), vector databases (Pinecone, Weaviate, Qdrant, Supabase), document loaders, memory systems, and tools into visual graphs. The resulting chatflow can be deployed as an API endpoint or embedded as a chat widget. This approach dramatically reduces the time to prototype and deploy AI applications compared to building from scratch in Python. Flowise is fully self-hostable on any infrastructure—a Docker container on a VPS, Kubernetes, or cloud provider. This is its most important characteristic for EU data sovereignty: when self-hosted on EU infrastructure, all data—documents, conversations, embeddings—stays within the organisation's own environment. The LLM provider can also be a local model (via Ollama) or an EU-sovereign inference API (Nebius, OVHcloud, Scaleway), making it possible to build fully EU-sovereign AI applications. Flowise also offers a cloud-hosted version (Flowise Cloud) for teams that prefer managed infrastructure, hosted in the US. For EU sovereignty, the self-hosted path is strongly preferred. The project has gained significant open-source adoption and the maintainers have formed a company around the cloud and enterprise offering, but the core open-source tool remains freely available.