At the Inter-American Development Bank, I built and maintained an evidence extraction, coding, and quality-control pipeline, including the codebook and validation rules used by the team. I also delivered systematic reviews and meta-analyses supporting education policy decisions across Latin America and the Caribbean.
At the University of Zurich’s Zurich Center for Neuroeconomics, I simulated decision-making-under-risk models in Python and optimised the pipeline. I also implemented computer vision models to generate saliency maps over image stimuli.
I built a scoring and classification model using the INDEC household survey, comparing ten algorithms and selecting a model based on false-negative cost rather than accuracy. In a separate project, I published an interactive dashboard with a geographic map and dynamic tables.
My work also includes designing and analysing three randomized experiments with more than 600 participants, as well as teaching behavioural economics and game theory at Universidad de San Andrés.
