RAG & AI automation / Workflow implementation

Give your knowledge a conversation.

A document ingestion pipeline and retrieval-powered chatbot, connected through Pinecone.

Project: Knowledge, connected.
Focus: RAG & AI automation
Technology: n8n / Pinecone / OpenAI / OpenRouter

RAG architecture showing Google Drive document ingestion into Pinecone and a chat agent with retrieval tools — Simplified diagram of the repository workflow. Execution requires your own service credentials and connected data sources. Open full preview ↗

THE THINKING BEHIND THE BUILD

The challenge.

Useful information lives inside documents. A conversational interface needs a dependable way to find the right context before preparing an answer.

The approach.

The workflow separates document ingestion from chat. New Google Drive files are downloaded, split and embedded into Pinecone. A chat-triggered agent uses a separate Pinecone retrieval tool and an OpenRouter model.

INSIDE THE PROJECT

Built around the details.

Document ingestion

Google Drive trigger and download nodes feed a loader, recursive text splitter and OpenAI embeddings.

Retrieval as a tool

The chat agent can query a Pinecone vector store for context from indexed documents.

Two connected paths

Ingestion and conversation are separate workflows within the same export, making the architecture easier to inspect.

HAVE A SIMILAR CHALLENGE?

Let’s build your next step.

Discuss a project

NEXT PROJECTDr Haroon Dental Clinic