AI support agent (n8n + PDF RAG)

Waybox - 2024-2025

Overview

AI agent connected to an internal PDF knowledge base to speed up support and improve answer quality.

Design and delivery of an n8n-orchestrated RAG pipeline: PDF preprocessing, cleaning, semantic chunking, embeddings, and vector indexing. The agent generates contextual, sourced answers (citations) with a confidence threshold and automatic human escalation when uncertainty is high. Goal: reduce response time, standardize quality, and improve knowledge capitalization.

Gallery

Tech

• Node.js
• n8n
• Qdrant/FAISS
• Python
• OpenAI/Ollama
• Docker

Problem & role

Design & implementation

Challenges

• Heterogeneous PDF quality (OCR, layouts, noise)
• Need for sourced answers and strict uncertainty handling
• Reduce resolution time without degrading quality

Solutions & impact

Solutions

• Cleaning pipeline + OCR when needed, normalization, and detection of useful sections
• Semantic chunking + citations + confidence threshold
• Automated human fallback and improvement loop via support feedback

Impact

• More consistent answers thanks to internalized documentation
• Estimate: response time for simple requests divided by ~2
• Estimate: level-2 escalations reduced by ~20-40%

Repo