Portfolio Case Study

Supplier data, ready for publication

The workflow combines scraping, data normalization and AI-assisted writing to automate the repetitive parts of product creation while keeping a final human validation step.

Make.com Google Sheets Web Scraping JSON ChatGPT
Video Workflow Demo Full process walkthrough

A future walkthrough will show the complete journey from the Google Sheets input to the final product review.

The Challenge

Creating product pages manually was slow and inconsistent.

Each product required manual supplier research, data extraction, image and document collection, copywriting, technical formatting and SEO preparation.

Repeating these tasks across different supplier websites made the process difficult to scale and increased the risk of missing or inconsistent information.

  • Different structures for every supplier website
  • Repeated copying and formatting work
  • Inconsistent product descriptions and attributes
  • Manual category and breadcrumb decisions
  • Time-consuming media and SEO preparation
Place a visual of the manual process, source files or repetitive tasks here.

The Workflow

A clear six-step process from input to validation.

Step 01 / 06

Product Input

The team enters only the essential product information in a simple Google Sheets interface.

Place the workflow overview or a Make.com scenario screenshot here.

Data Structure

Different supplier formats become one consistent product model.

Standardized information

Every supplier exposes information differently. The automation maps the collected fields into a common JSON structure so downstream modules always receive predictable data.

  • Product name and description
  • Images and technical documents
  • Key specifications
  • Warranty and certifications

Controlled content output

The writing module follows a fixed content framework instead of generating unrestricted copy.

  • Clear product introduction
  • Main customer benefits
  • Technical characteristics
  • Installation or usage guidance when relevant
Place a structured data example or generated product description here.

Quality Controls

Automation accelerates production without removing oversight.

Normalize supplier data before it is used
Match standardized technical attributes
Validate category and breadcrumb placement
Generate SEO titles and descriptions from rules
Keep a final human review before publication
Place the category matching, breadcrumb or SEO validation screen here.

Measured Impact

2 hours reduced to about 20 minutes.

Before 2 hours Manual creation
After About 20 min Generation and review
Batch size
6

Product pages per measured batch.

Automated generation
5 min

Automated generation.

Review phase
15 min

Human quality review.

Time saved
1h 30

Saved per six-product batch.

Place the final product sheet or publication-ready result here.

My Contribution

Designing the logic behind a reliable end-to-end workflow.

Workflow Design

Mapped the manual process and translated it into clear automation stages.

100%

Scraping and Mapping

Structured supplier-specific collection logic and normalized the resulting data.

95%

AI Content Rules

Defined the prompts, content structure and checks used for product copy.

90%

Quality Assurance

Built consistency checks and retained human validation before publication.

85%

Conclusion

A faster process with more consistent product information.

This project shows how scraping, structured data processing and AI can remove repetitive work from product creation while preserving quality control.

The most important result is not only speed. The shared data model, controlled writing rules and validation step make the process easier to maintain, scale and trust.