Services

We start with what actually eats your time: copying data between systems, manual spreadsheets, monthly reports assembled from five sources, handling files from metering devices or third-party systems.

The result is not “a script on somebody’s desktop”. It is a process with logging, error handling and documentation — one that keeps working when you are on holiday.

Typical work: measurement and billing file processing, automated reports (Excel, Word, PDF), data synchronisation between systems, RPA for desktop applications that have no API.

When a spreadsheet stops being enough but a full ERP is a cannon aimed at a fly, a small, well-designed application wins: one screen, one job, no training needed.

Built on Python (Django) and React — a stable, well-documented stack that will still be maintainable in five years, including by someone other than me.

You rarely get a full set of clean APIs. It is usually a mix: a REST API from one vendor, a CSV export from another, e-mail attachments from a third, and a desktop application with no interface at all.

My job is to tie that into one predictable flow and document it so it can be maintained.

Automating a bad process gives you a bad process that runs quickly. So every project starts with talking to the people who actually do the work, and drawing the process out step by step.

Very often it turns out two steps are unnecessary, one can be solved with a setting in a system you already own, and only the rest needs automating. I say so openly — even when it means less work for me.

Excel, Word, PDF, database, dashboard — the format should be whatever suits the reader, not whatever suits the developer.

I cover the whole path: fetching the data, validating and cleaning it, running the calculations, generating the document and delivering it where it needs to go.