Julien Gibert

I took over the redesign of an existing n8n workflow to optimize the internal reporting process. The main goal was to make meeting minutes generation more reliable while ensuring rigorous document management.

Details of the implementation:

Audit and debugging of existing n8n scenarios.

Implementation of strict logic for adherence to file naming conventions (dynamic naming).

Automation of the archiving and distribution of summary notes.

Time savings and elimination of manual data entry errors.

I design and deploy a complete automation for creating digital product sheets on WordPress using n8n and OpenAI, ideal for online stores looking to save time, reduce errors, and industrialize their digital content production.

What I accomplished:

I developed an advanced workflow combining n8n, OpenAI, WordPress API modules, and third-party services (image conversions, PDF splitting) to enable:

Mission Objective:

Industrialize the creation of product sheets for a WordPress WooCommerce e-commerce site specializing in digital product sales, while ensuring:

• a reliable and scalable process

• 10x faster content production

• consistent and optimized SEO quality

• reduction of human errors in publishing

Technologies Mastered:

• n8n (complex workflows, advanced expressions, HTTP Request, Webhooks, Queue mode)

• OpenAI (content generation, JSON structuring, automatic validation)

• WordPress / WooCommerce REST API

• WordPress Media Library + downloadable files

• PDF → PNG conversions

• SEO metadata management (Yoast)

• Workflow optimization and e-commerce automation

Result:

A turnkey, robust, and fully automated solution that allows publishing a complete product sheet in seconds, without human intervention.

I built an n8n workflow that reads targeted incoming emails, automatically extracts customer information with an LLM, calculates distances (Google Maps Distance Matrix), and populates a prospect Google Sheet. Result: zero copy-pasting, clean and actionable data in 1 minute.

In the moving business, requests come in via email and are manually re-entered. I built an n8n workflow that captures, structures, enriches, and automatically follows up with prospects — automatically.

1. Filtered Gmail Trigger (polling 1 min): from:[email protected] OR from:[email protected].

2. LLM Extraction (OpenAI GPT-4.1-mini) with strict JSON: gender, last name, first name, phone, email, departure/arrival addresses, volume (m³), floor/difficulties, comments, moving date (DD/MM/YYYY). Missing fields → "".

3. Geolocation Enrichment (Google Distance Matrix metric): departure↔arrival, depot(Vichy)↔departure, arrival↔depot to estimate costs/logistics.

4. Google Sheets (append): normalization (correct case for Name/First Name, lowercase for email), columns ready for CRM/BI.

5. Automatic sales follow-up: 3 sequential emails per prospect (timed/spaced) to maximize response rate, personalized with extracted data.

Results.

• -90% lead processing time, clean and consistent data.

• Responsiveness + preliminary quote via calculated kilometers.

• Centralized tracking (Sheet/CRM) + automatic follow-ups to increase conversion.

Stack & Best Practices.

• n8n (no-code/low-code), Gmail, OpenAI, Google Maps, Google Sheets.

• Secure API keys in n8n Credentials (never in plain text).

• Defensive prompting (temperature 0), stable column mapping, executable logs.