UWork in the fraud area in the detection of user profiles intended to perform fraudulent actions in the application. I do this detection through the development of machine learning models and exploratory analysis of data and interconnections between users. Tools: Python, Snowflake, PowerBI.

Reduce by 42% the monthly loss due to abuse of discounts.

Supported the research agenda in data pre-processing and development of Bayesian predictive models to forecast nursing home hours of duty and optimize shift assignment.

Development of a module in Python for the prediction of the demand for nurses through Bayesian models of time series.

Clean Transport Project: Estimated polluting gas emissions in cargo vehicles, participated in the development of the web application for the presentation of the estimates, and led a team of 3 developers. Vigia Project: Created indicators and algorithms for the prediction of irregularities and inefficiencies in public procurement in Bogotá, and I participated in the development of the web application to optimize public supervision.