Omnilog

Uruguay

E-commerce

We migrated the web scraping process for 5 supermarkets in 3 weeks using AI, achieving results that were 90% faster following a massive overhaul of their platforms. In just 7 days, we restored this critical operation to production.

Python

AI-Assisted Development

7 Days

Scripts in Production

+90%

Faster

21 Days

Full Migration

How do you

help you?

Python

AI-Assisted Development

Challenge

In early 2026, several supermarkets switched their e-commerce platforms at the same time, forcing an urgent migration of all the scraping scripts. The scripts functioned as a Windows installer that saved results locally, and stopping the scraping process meant losing data critical to the business. The team was part-time, and the time available was minimal. The challenge wasn’t technical in itself—it was getting it done quickly.

Solution

Using spec-driven development and Claude Code, we analyzed each supermarket’s new e-commerce platforms, defined the scope, and developed the Python scripts. In parallel with the migration, we implemented structural improvements: dynamic parallelism, logging to local files, and a shared observability strategy. All of this was accomplished with one senior developer working part-time.

What was the

result?

Within a week, the client already had its main scripts running in production.

Something that would have taken three times as long before. Parallelism improvements made it possible to obtain results 90% faster, and shared observability significantly reduced friction when reporting and resolving errors. Five supermarkets migrated in three weeks, with no downtime.