AI Agent for Appian Tasks: Ollama LLM + Postgres Memory

This n8n workflow creates a conversational AI agent that interacts with Appian tasks via a chat interface. It uses a local Ollama Qwen2.5:7b LLM for reasoning, Postgres for persistent conversation memory keyed by sessionId, and custom tools to fetch user tasks or create new ones (with title, description, hours, cost). The AI follows system rules as a helpful assistant, handling date formatting and iteration limits. Key components include a webhook or chat trigger for user input, field normaliza