Is It Feasible to Use Xano Embeddings with n8n AI Agents?

Based on the official documentation, n8n provides built-in support for Supabase (as well as Pinecone and in-memory vector stores) to be used as vector databases with AI Agent workflows. There isn’t any documented or built-in node for Xano embeddings at this time.

If you want to leverage Xano's vector embeddings with n8n AI Agents, here are some potential approaches:

• Custom Integration via HTTP Requests:
  You can use an HTTP Request node in n8n to interact with Xano’s API endpoints. This would let you push or retrieve your vector data from Xano. You’d need to manage authentication and the data transformation yourself.

• Building a Custom Node:
  If you plan to use Xano extensively for embeddings, you might consider developing a custom node. This would encapsulate your HTTP interactions in a way that feels native to n8n workflows.

• Hybrid Strategy:
  Continue using Supabase (or another supported vector store) for the core vector search functionality while using Xano as your back-end for relational data. This keeps things simple as long as duplication is acceptable.

At the moment, there are no community-provided examples or official guides showing integration of Xano embeddings directly with n8n AI Agents. It might be worth experimenting with the HTTP Request node approach or checking the n8n community forums for any emerging solutions.

Let me know if you need more details on how to set up custom HTTP interactions in n8n!