Jorge Carrasco
Data Analysis. Python. SQL.
Projects
E-commerce sales financial report
Sales analysis project for an E-commerce sales dataset using Excel, Power Query and Power Pivot to create a financial report.
Dashboard: USA Education 2020 vs 2021
Interactive dashboard about the education landscape in the US for the years 2020 and 2021. Developed with python, pandas and streamlit.
Density-Based Clustering for Automatic Detection of Criminal Hot Spots in Mexico City
Study of the Distribution of Crimes in Mexico City between 2019 and 2021 using the OPTICS clustering algorithm to automatically detect areas with high crime incidence.
2022
See more
Experience
Deep Axiom
Data Scientist
- For a project in the livestock sector, I developed and implemented computer vision models to detect irregularities in the behavior of personnel at access and control points of the facilities. I developed these models using the YOLOv5 architecture, python and the ultralytics package. These models increased the effectiveness of the client's security system.
- Designed and developed a monitoring system for the livestock process to maximize income (sales per kilogram) by reducing losses due to irregularities in the breeding, reproduction, transportation, feeding, medical history and administrative processes. The solution I implemented with python, streamlit and plotly provided key insights into the metrics, performance and irregularities of each part of the livestock process.
04/2023 - 08/2023
Centro Oncológico Integral – Hospital Ángeles
Data Science Intern
- Designed and developed a computer vision model to automate 3D reconstruction of vital organs from CT scans. I implemented the model based on the U-Net architecture using python, tensorflow and keras, obtaining DSC and IoU similarity scores greater than 90%. The development is being used by the Hospital and marketed in the private healthcare sector.
- Developed multiple computer vision models for early assisted detection of cancer from CT scans. I processed the images and data with python, torchvision and scikit-image and implemented the models with python, tensorflow, keras and ultralytics.
- Worked in close communication with oncologists and software engineers to develop models that met medical requirements and were suitable for deployment by the software team.
01/2023 - 04/2023
Facultad de Ciencias Físico Matemáticas
Research assistant
- Designed a population dynamics model between an insect pest species, the use of biological control (predatory species) and the application of chemical control in almond orchards to propose pest management strategies when there is resistance to chemical control. I developed the model based on the reported biological behaviors of the pest species and biological control under the effects of chemical control. Additionally, I implemented and adjusted the model to experimental data using python, scipy and numpy.
01/2022 - 12/2022
XIV Congreso Mexicano de Inteligencia Artificial
XIV Mexican Congress of Artificial Intelligence
- Study of crime distribution in Mexico City between 2019 – 2021 using the OPTICS clustering algorithm to automatically detect areas with high crime incidence. I developed the research project using python, sci-kit learn, scipy, geopandas, pandas and matplotlib.
- Participated in the XIV Mexican Congress of Artificial Intelligence – COMIA (May 2022)
- Publication of research paper in the journal Research in computing Science (August 2022)
05/2022 - 08/2022
Applied Mathematics Bachelor's degree
Benemérita Universidad Autónoma de Puebla
- Cummulative percent average: 96.0
- Elective subjects related to data science
08/2018 - 06/2023