Andres Granados

I maintained a web scraping tool based on Scrapy in an architecture build of Docker containers to host the scraper, the headless browser and other applications for ETL of metadata that communicate through RESTful APIs using FastAPI. I measure the quality of the AI-generated metadata compared to those validated by editors. I implemented a generalization of the Recall and F1-scores that uses WordNet and GermaNet to compute the similarity between words and sentences generated by AI summarization.

I developed and deployed machine learning methods for Natural Language Processing in: sentiment analysis by clustering the customer service requests. Computer vision: facial identification, medical imaging diagnostics and monitoring areas to prevent the spread of a virus. Computer simulation of the flux of people in public areas to estimate the risk of spreading a virus. Big data for telecommunication industry based on Hadoop ecosystem technologies.

Development and deployment of web services and web applications hosting machine learning and artificial intelligence applications.

We made a predictor of drop-out in school students. We developed a user interface for visualize the percentage of students in high risk of drop-out in a given school and timelapse.

I taught the students of the Introduction to Machine Learning course the content of scientific programming with python for machine learning tasks.

Theoretical Physics, Astrophysics, Data Analysis, Data visualization, Numerical methods and scientific programming