aıpro

← More pros

Day job · Data collection

Crawlers that feed on the whole web

Her current work is web scraping at production scale: building and maintaining Python crawlers that collect and process data from various websites, day after day. It is unglamorous, essential work. Data products are only as good as the crawlers feeding them, and hers are the kind that keep running.

Around the crawlers sits the supporting craft: backend systems, API integrations, and automation, the plumbing that turns raw collected pages into data a product can use.

Work build · GPT API

An email categorization tool

As a work task, she built an email categorization tool using Python integrated with the GPT API: incoming email sorted by a model instead of by hand. It is her clearest AI build so far, a working tool rather than an experiment.

There is no public demo yet, and she is interested in preparing one. Her AI toolkit in the meantime is practical: ChatGPT and Devin AI, used to debug and improve code, understand errors, and solve programming problems, with the hardest problems usually in the scraping work itself.

In the open · GitHub

The code, in public

Screenshot: Pinky Torres's GitHub profile
Live screenshot: her GitHub, where the pinned repositories are the scraping and parsing tools themselves. Tap the picture to visit.

Her GitHub holds 14 public repositories, and the pinned ones are the day job in miniature: a data scraping project, an exported data parser, an admin panel, and her align-python-library, kept in three versions as it evolved.

It is a specialist's paper trail: parsers, libraries, and crawlers, versioned in public, going back years. What she says she does is what the repositories show.