I’m trying to build a Q&A flow in n8n that uses Azure OpenAI Embeddings together with a Weaviate Vector Store retriever. The goal is:
-
Trigger: When a chat message is received
-
Chain: Pass query → Azure OpenAI Embeddings → Weaviate Vector Store → return top-k results
The issue
I keep getting the error:
400 Unsupported data type
This only happens in the Embeddings Azure OpenAI node. The Chat model works fine.
The Weaviate Vector Store node ask for a collection which is “Article”.
In Weaviate, article contains:
-
Named vector: te_3_large
-
Properties: title, lead, body, userNeed, etc.
The question is how do I retrieve the data from this vector database with the right data type?
