From visual flow design to multi-agent orchestration, CX-Builder provides the complete toolkit for enterprise AI development.
Design AI workflows on an intuitive canvas. Three builder modes — Assistant, Chatflow, and Agentflow — scale from simple chatbots to complex multi-agent systems.
Create teams of AI agents with supervisor-worker patterns. Agents collaborate, delegate tasks, and solve complex problems through hierarchical coordination.
Connect 42+ data sources — PDFs, databases, APIs, cloud storage — and build retrieval-augmented generation pipelines with pgvector-powered semantic search.
OpenAI, Anthropic Claude, Google Gemini, Mistral, Groq, AWS Bedrock, and local models via Ollama. Switch providers without rewriting workflows.
Keep humans in control with approval gates, review checkpoints, and input requests embedded directly into automated agent workflows.
Role-based access control, SSO, JWT authentication, encrypted credentials, rate limiting, and full audit trails. Deploy on-premises or in your cloud.
Execution traces, visual debugging, and integrations with LangFuse, LangSmith, and other analytics platforms for complete pipeline visibility.
Deploy your agents anywhere with a customizable chat widget, REST APIs, and TypeScript / Python SDKs for seamless integration.
Customer support bots, RAG pipelines, SQL agents, research assistants, and multi-agent teams — production-tested templates to start building instantly.
Just Docker — nothing else required. No git clone, no manual config. The installer generates all secrets automatically and starts the full stack.
curl -fsSL https://raw.githubusercontent.com/dtsoden/CX-Builder/master/install.sh | bash
irm https://raw.githubusercontent.com/dtsoden/CX-Builder/master/install.ps1 | iex
A containerized stack with pgvector-powered semantic search, multi-platform support, and enterprise-grade security — ready for any scale.
Visual Builder + REST API
Vector + Relational DB
LLMs, Tools & Data
Open source. Self-hosted. Deploy the full platform with a single command and start creating AI agents in minutes.