Ania Boutarene

Two master's degrees, one in Algeria, focused on AI and systems thinking, and the other at Stockholm University's Computer and Systems Sciences. The turning point was a research internship at KTH, where I built deep-learning models for heat-load forecasting. That's where it clicked: clean, structured input is where intelligence starts.

The intellectual puzzle. How fields map across systems and how patterns in raw data point to something meaningful. And then there's the craft of building it from scratch. My dad was a handyman; growing up, I'd assemble furniture just for the fun of it. That instinct never really left.

Routine lookups went from slow to instant, and each round of tuning improved accuracy substantially. By the final configuration, the question had stopped being "does this work?" and started being "what do we do next?"

I didn't expect to enjoy the finance domain as much as I did. Suddenly I got genuinely curious about invoice ageing, ERP structures, what account managers need day-to-day. Once you understand what people need, technical decisions stop feeling abstract. That's one of the quiet privileges of consulting: every project is also an education.

Before asking what AI can do, you must ask whether the data underneath it deserves to be trusted. On this project, NULLs, inconsistent status codes, and sign normalisation issues would have caused the agent to return wrong answers if left unaddressed. That work is rarely glamorous, but it makes everything else possible.