Try the work, not just read about it
Small demos built from patterns in my real projects. Each one is labelled: live means real code running in your browser on sample data; simulation means a faithful model of a system that would be unsafe or costly to run publicly. Nothing here scrapes third-party websites.
Specialist problems
Simulations of the hard parts of real projects: awkward sources and the infrastructure behind them.
Difficult-source lab
Run a quick script and an engineered collector against the same awkward website, and see why "it ran without errors" is not "the data is complete".
Shows: JavaScript rendering, stateful paging, expired sessions, bot checks, rate limits, incomplete records and completeness checks.
Open the demoMobile proxy farm
Drive the rotation logic of a real Raspberry Pi and LTE-modem proxy farm: traffic-driven rotation, the keep-one-online rule, and a modem that fails to come back.
Shows: Rotation state machine, HAProxy runtime control, health checks, failure handling and the trade-offs behind them.
Open the demoApproachable tools
Smaller live tools and simulations you can use on sample data, or on your own file.
Page to data
Point at a sample product catalogue and turn three messy pages into a clean table, then export CSV, JSON or an Excel-compatible .xls file.
Shows: Field mapping, pagination, normalising prices, dates and ratings, deduplication, and export.
Open the demoResilient job runner
Run a 240-record extraction job, throw failures at it, kill it halfway, and watch it resume from its checkpoint.
Shows: Retries with backoff, rate limits, checkpoints, resuming without duplicates, and honest final reports.
Open the demoCSV cleaner
Drop in a messy CSV. Trim, fix casing, normalise phones, emails and dates, remove duplicates, and download the result.
Shows: The data-cleaning layer every pipeline needs, with a visible diff of what changed. Your file never leaves your browser.
Open the demoAutomation estimator
Estimate what a manual task costs per year and when automating it would pay back. Every assumption is visible.
Shows: Whether automation is worth it before anyone writes code.
Open the demoWant one of these built for your real data?
Describe it in your own words. I reply within two working days with whether it looks doable, the main risks and a sensible first step. The first conversation is free.