Run LLM assistants over company data with RAG in n8n
A commenter used multiple LLM assistants in an n8n project at their company, handling AutoCAD Forge API calls, parsing, embedding and RAG across company data. Task: Suggested implementation: Configure n8n workflows to deploy multiple LLM assistants for AutoCAD Forge API calls and company data processing via parsing, embeddings, and RAG. The system aims to simplify initial setup compared to Python projects, with human oversight required to monitor request routing and data consistency.; tools: n8n. Human role: Source evidence is unavailable regarding specific human roles or manual checks in the described workflow.; limitations: Prerequisites include n8n, LLM API keys, and AutoCAD Forge access. Constraints: n8n offers easier startup than Python, but Python may handle specific API tasks more robustly. Risks include inconsistent responses, difficult RAG/parsing setup, and routing errors to wrong databases. Source reliability is low (HTTP 403)..
Tools
Published on Agent Work Atlas:
Overview
A commenter used multiple LLM assistants in an n8n project at their company, handling AutoCAD Forge API calls, parsing, embedding and RAG across company data.
Workflow
Suggested implementation: Configure n8n workflows to deploy multiple LLM assistants for AutoCAD Forge API calls and company data processing via parsing, embeddings, and RAG. The system aims to simplify initial setup compared to Python projects, with human oversight required to monitor request routing and data consistency.
Reported result
Actual outcomes are unknown due to source unavailability. Proposed acceptance check: Measure the percentage of correctly routed requests versus misrouted ones, and evaluate response consistency across assistants. Expect potential issues with API integration complexity and information spread between agents.
Human role
Source evidence is unavailable regarding specific human roles or manual checks in the described workflow.
Limitations
Prerequisites include n8n, LLM API keys, and AutoCAD Forge access. Constraints: n8n offers easier startup than Python, but Python may handle specific API tasks more robustly. Risks include inconsistent responses, difficult RAG/parsing setup, and routing errors to wrong databases. Source reliability is low (HTTP 403).
Try it
Act as an n8n workflow engineer. I need to build a pilot RAG system for [replaceable: company document set] using [replaceable: LLM API key]. Steps: 1. Create an n8n workflow with multiple LLM assistants. 2. Implement parsing and embedding for [replaceable: source files]. 3. Configure RAG retrieval and AutoCAD Forge API integration. 4. Set up Supabase for data storage. 5. Run a dry run with 10 sample queries. 6. Manually verify routing accuracy and response consistency. 7. Report metrics: % correct routing, % consistent responses. If routing errors exceed 10%, pause and adjust agent logic. Do not publish or spend credits until I confirm the dry run results.
Source reports are not independently verified. Publication date is the catalog publication date, not the original article date.