Petra is an applied mathematician turned AI engineer. She holds a master’s degree in applied mathematics from the University of Zagreb, where her thesis modelled the climate effects of an asteroid impact through numerical simulation of complex physical systems.
At SciPly she runs the company, leads our AI work and builds our own products from scratch, turning real pain points into working prototypes, predictive tools and automated workflows. Before that she spent over three years at A1 Croatia, growing from data analyst to senior data scientist: she built a daily cash-flow forecasting tool for the finance team, prototyped offer recommendation, credit scoring and fraud detection models, and replaced repetitive manual work with automated Python and SQL pipelines. Earlier, she interned in banking supervision at the European Central Bank in Frankfurt.
What she works on
- AI products: LLM-powered backends in Python and FastAPI, built on the Gemini API and Google Cloud, from first prototype to production.
- Machine learning and forecasting: time-series models, classification and recommendation systems with Prophet, XGBoost and scikit-learn.
- Business translation: working directly with stakeholders to scope what they need and turn it into practical, automated solutions.
Projects and community
She founded MyEndoDiary, an AI-assisted symptom diary for people living with endometriosis and adenomyosis, and leads it end to end, from user discovery to the Flutter app and its cloud backend. She is a community lead at Google Developer Group Zagreb, speaks about machine learning at tech events, and has mentored young women in STEM through STEMFemme.