Operations & Systems
Open WebUI Agent with Web Search
This workflow creates an Open WebUI compatible AI agent with web search capabilities. It uses Postgres for chat memory, Ollama for the chat model, and SearXNG for advanced web searching, providing an entirely local RAG solution.
Built · ~9.5 hours saved per week
Users who want to interact with an AI assistant that can browse the web and remember past conversations.
This workflow provides an AI chatbot that answers questions using web searches. It remembers past conversations and can get information from the web to give you updated answers.
1. Set up an n8n instance. 2. Install the required nodes. 3. Connect to a PostgreSQL database for chat memory. 4. Configure an Ollama API key for the AI models. 5. Set up Open WebUI authentication. 6. Deploy as a webhook.
n8n, PostgreSQL, Ollama, SearXNG (or similar self-hosted search engine).
You can change the AI model, modify the system message guiding the AI, and adjust how many search results the AI reviews.
Key Technologies: n8n,PostgreSQL,Ollama,SearXNG,Langchain
Automation Value: AI Chatbot Interaction Contextual Conversations Web Search for Information Information Extraction
Best Practices: Modular Design with Sub-workflows Error Handling Clear System Messages for AI Context Management Local AI/Search Integration Input Validation/Routing
Tags: Chatbot, Web Scraping, Webhook Trigger, Admin, Local Files, AI Assistant
Integrations: Postgres, AI Agent, HTTP, Ollama, Webhook