Skip to main content

Flowise

Visual AI agent builder

Flowise provides a drag-and-drop interface for building AI agents with LangChain components.

Overview​

  • URL: http://localhost:5003
  • Image: flowiseai/flowise:latest
  • Purpose: Visual agent creation and flow building
  • Status: Always enabled

Features​

  • Visual Editor: Drag-and-drop flow builder
  • LangChain Integration: Pre-built components
  • Multiple LLMs: OpenAI, Anthropic, Google AI
  • Memory Management: Conversation persistence
  • Vector Databases: Document embeddings
  • Custom Tools: API integrations

Configuration​

Required​

FLOWISE_PORT=5003
FLOWISE_PASSWORD=seiling123

Optional​

OPENAI_API_KEY=sk-proj-your_key
ANTHROPIC_API_KEY=your_claude_key

Usage​

First Time Setup​

  1. Open http://localhost:5003
  2. Login: admin / seiling123
  3. Create new chatflow

Basic Agent Creation​

  1. Add LLM Node: ChatOpenAI or similar
  2. Add Memory: Buffer or Redis memory
  3. Add Tools: Custom tools or APIs
  4. Connect Flow: Link components
  5. Test Agent: Try sample queries

Available Components​

  • LLMs: OpenAI, Anthropic, Google, local models
  • Memory: Buffer, Redis, PostgreSQL
  • Vector Stores: Qdrant, Pinecone, Chroma
  • Tools: HTTP requests, databases, custom functions
  • Chains: Sequential, conversation, RAG

Integration​

With Seiling Services​

  • Qdrant: Vector storage for RAG
  • Redis: Memory and caching
  • PostgreSQL: Data persistence
  • Sei MCP: Blockchain tools

Common Patterns​

  • RAG Chatbot: Documents + LLM + memory
  • Tool Agent: LLM + custom tools + memory
  • Research Assistant: Web search + analysis + summary

Troubleshooting​

Can't Access Interface​

# Check service
docker logs seiling-flowise

# Verify port
curl http://localhost:5003

Flow Execution Errors​

  • Check API keys in settings
  • Verify component connections
  • Test individual nodes
  • Review execution logs

Flowise enables visual AI agent development. Perfect for building complex conversational agents without coding.