Hr ops · Production

XIBIX Solutions reduces HR inquiry load by over 50% with n8n-powered RAG bot

The problem

XIBIX needed to move from isolated AI proofs of concept to repeatable internal services accessible to both technical and non-technical staff, while keeping workflows under their own control. Separately, a small HR team of three to four people was overwhelmed by repeated employee questions about internal documentation.

First attempt

Earlier experiments with Make and Power Automate introduced tradeoffs around control, usability, and operating model. Employees also failed to adopt AI tools that required connecting to a separate interface outside their existing chat window.

Workflow diagram · grounded in source
1
Employee asks HR question
trigger
“Employees needed quick answers like where to find vacation information or how to request HR paperwork”
2
API routes to tagged n8n workflow
routing
“That API inspects n8n workflows tagged for their internal "XGPT" and exposes them as selectable models”
3
HR content ingested from Confluence
integration
“n8n pulls HR content from Confluence, prepares embeddings, and uses a hosted vector database in Azure”
4
RAG agent answers question
ai_action
“n8n then runs the retriever and agent logic that answers employee questions based on the company's internal documentation”
Reported outcome

XIBIX reduced repetitive HR inquiries by more than 50% and HR saves more than 50% of the time previously spent re-answering the same questions, freeing the team to focus on higher-value work.

Reported metrics
repetitive HR inquiries reducedmore than 50%
HR time saved on repeated questionsmore than 50%
Reported stack
n8nConfluenceAzureOpen WebUIMicrosoft Teams
Source
https://n8n.io/case-studies/xibix-solutions/
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

XIBIX reduced repetitive HR inquiries by more than 50% and HR saves more than 50% of the time previously spent re-answering the same questions, freeing the team to focus on higher-value work.

What tools did this team use?

n8n, Confluence, Azure, Open WebUI, Microsoft Teams.

What results were reported?

repetitive HR inquiries reduced: more than 50%; HR time saved on repeated questions: more than 50% (source-reported, not independently verified).

What failed first in this deployment?

Earlier experiments with Make and Power Automate introduced tradeoffs around control, usability, and operating model.

How is this hr ops AI workflow structured?

Employee asks HR question → API routes to tagged n8n workflow → HR content ingested from Confluence → RAG agent answers question.