Travel Planning AI Agent: Couchbase Vector Search, Gemini & OpenAI

This workflow template features two interconnected workflows for building an intelligent travel planning agent. The Data Ingestion workflow uses a Webhook to receive points of interest data, generates embeddings with OpenAI, splits text recursively, and stores vectors in Couchbase using the community Couchbase Vector node. This sets up a vector database for semantic search on travel destinations.

The Chat Application workflow powers a conversational interface via Chat Trigger, employing an AI Tools Agent with Gemini 2.0 Flash as the chat model for natural responses, Simple Memory for context retention, and Couchbase Vector Search as a tool for retrieving relevant travel info based on user queries. OpenAI embeddings ensure accurate similarity matching.

Benefits include rapid deployment of RAG-based AI agents, scalable vector search without custom coding, and enhanced user experience through contextual chat. Ideal for travel apps, tourism sites, or learning Agentic AI with real-world vector stores. Saves weeks of development time on embeddings, storage, and agent orchestration.