Consulting
Four areas. They overlap, and most real work draws on more than one.
Data engineering
Moving data between systems that were not designed to talk, and making the result trustworthy: what a column means, what an empty value means, which side wins when two systems disagree. The interesting part of this work is rarely the pipeline. It is deciding what the data actually says, and writing that down where the next person will find it.
Business applications
The systems a company runs on rather than sells: intake, approval, records, reporting. Usually there is an obligation behind them and a process that grew by accident around it. The work is to find the rule the process is really serving, and build something that serves it deliberately.
Perl
Modern Perl, and the older code that pays the bills. If you have a working system nobody wants to touch, that is a normal situation and not a shameful one. It can be tested, documented and changed safely, and rewriting it is rarely the cheapest route. We also publish to CPAN, so this is not archaeology at a distance.
Skunkworks projects
A short, bounded project to answer a question before anybody commits a budget: can this be done, what would it cost, what breaks first. The deliverable is an answer you can act on, including the answer "not worth it", which is worth the same money and saves considerably more.
How engagements usually start
A conversation about what is actually in front of you, then a short written scope with a fixed price for the first piece of work. If the first piece shows the problem is different from the one you described, we say so and re-scope rather than proceed.