Help 🙏 me bind the issuance of links to the source in the RAG Agent, as recommendations

Questions
1

Help link the issuance of links to the source, as recommendations by the Agent. + Question on architecture and nodes

Friends, hello!:sunny:
I need help from the community in developing a workflow for YogaAssistant

I would appreciate any help and templates. The logic of the workflows resembles a researcher and a doctor. I’m a newbie and I’m stuck.:exploding_head:

We need to make an assistant for Yoga School that gives links (about 40 pages with courses) to relevant courses from the site after the user has reported their problem

Each link is a separate page dedicated to a health problem, for example, Breathing Problems or Hip Problems

My YogaAssistant should collect personal data from the user (Greating Agent), then receive a message from the user with information about diseases and what worries him and with clarifying questions (if required) collects the user’s anamnesis (Anamnes Agent), then gives a message about the user’s condition and recommendations for recovery, and also gives 2-3 links to site pages with relevant yoga courses.

Stack: N8N + Supabase + Telegram
RAG Multiagent

:question::question::question:

  1. How to correctly place links to pages or upload pages to Supabase, maybe make a table with a list of links and a brief description of the contents of each page? Please tell me how to do it correctly?

  2. How to implement the architecture of these 3 Agents more correctly? Consistently?

  3. I would be grateful if you write the sequence of nodes that can be contained in my workflow

:pray::pray::pray:
I really need help from an expert in N8N multi-agents RAG, I need to show it to the Customer tomorrow, and I’m stuck. I’ve collected several WorkFlows, including on DiF, but the logic and answers are not satisfactory.

Colleagues, I would be grateful for any useful advice, template or video.

Thanks in advance​:pray::pray::pray:

2

It looks like your topic is missing some important information. Could you provide the following if applicable.

  • n8n version:
  • Database (default: SQLite):
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app):
  • Operating system:
4

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.