From the course: Build with AI: Create Custom Chatbots with n8n
Unlock this course with a free trial
Join today to access over 24,700 courses taught by industry experts.
Creating the retrieval workflow in n8n
From the course: Build with AI: Create Custom Chatbots with n8n
Creating the retrieval workflow in n8n
- [Instructor] Now that we have a workflow that stores data in Pinecone, we also need a workflow that is actually retrieving data from Pinecone. So let's set this up now. Select +, Workflow, and create a new workflow here from scratch. So in this pane, what triggers this workflow, select When executed by another workflow. Just keep in mind the data ingestion right now runs every time we trigger it manually. The data retriever workflow, however, should be triggered every time the chatbot is actually looking for internal information. And this is what we'll later on add to our chatbot as a custom tool. Let's add a field here called query and make sure the type is set to String. Let's execute a step. You can see right now the query is null. For demo purposes, let's give this an actual value. Like, for example, sick days. This will emulate a search query that the chatbot later on is giving to our retrieval flow. Let's hit Save and close this window. Next, we'll add the Pinecone Vector…
Contents
-
-
-
-
(Locked)
Retrieval-augmented generation (RAG) in five minutes10m 36s
-
(Locked)
Setting up Pinecone2m 24s
-
(Locked)
Creating the embedding workflow in n8n12m 20s
-
(Locked)
Creating the retrieval workflow in n8n6m 1s
-
(Locked)
Step by step: Advanced chatbot workflow in n8n9m 2s
-
(Locked)
Milestone: Smart chatbot powered by your own docs3m 38s
-
(Locked)
-
-