At Omixom SRL, I work as a Full Stack Developer on an offline-first synchronization API and mobile app for service technicians. My work also includes time-series anomaly detection, a platform-data chatbot using OpenAI function calling, scheduled PDF and Excel reports, and a public air-quality dashboard for Río Cuarto.
For my final degree project, I reimplemented a biologically plausible probabilistic-inference model in JAX within Scikit-NeuroMSI, validating numerical equivalence with the original code and reproducing stimulus-modulated variability. I’ve also contributed a vectorized implementation of the Reference Ideal Method, with tests and documentation, to scikit-criteria (QuatroPe).
