Élio Pereira
Geospatial Data Scientist
Contact
Skills
Languages
English
(Fluent)
Portuguese
(Native)
Interests
Experience
Geospatial Data Scientist
- Created Python Machine Learning pipelines for downscaling land surface temperatures in Denmark and air pollutant concentrations in the Iberian Peninsula;
- Fetched and processed remote sensing data from a multitude of satellites (Sentinel-2, Sentinel-3, Sentinel-5P and Landsat 8/9)
- Thoroughly analysed observational and reanalysis data such as E-OBS, CAMS' European and Global Reanalysis as well as CERRA;
- Mastered Earth Observation Python packages such as geopandas, xarray, rioxarray, rasterio, folium and osmnx;
Simulation Engineer
- Developed a multiphysics simulation Python code for computing air flow rates, pressures, temperatures and heat loads in the wind turbine components;
- Produced respective Sphinx documentation;
- Created extensive regression models for physical variables of the wind turbine components, validated against experimental data.
Oct 2021 - Oct 2022
Porto, Portugal
Simulation Trainee
- Consultancy on Physics and Mechanical Design topics to other teams;
- Performed CFD simulations of internal flows in several components of Vestas' wind turbines;
- Worked under a Scrum framework, managed through Azure DevOps and Jira.
Education
Aerospace Engineering (MS)
- Grade: 17 (/20);
- Master's thesis: State-to-state Modelling of High-speed Nitrogen Shocked Flows , 19 (/20).
Sep 2013 - Jul 2016
Lisbon, Portugal
Projects
July 2024 - Aug 2024
- Description: construction of a CNN model for classifying skin lesions - benign and malign.
Stock Market Prediction - Dow Jones' Index
July 2024
- Description: construction of an LSTM to predict weekly stock prices of Dow Jones' 30 prominent companies.
Chemometrics - A Study on PCR and PLSR applied to Absorption Spectra
June 2024
- Description: thorough mathematical derivation of PCR and PLSR and their usage in a Chemometrics problem.
Apr 2024
- Description: detection of personally identifiable information (PII) in student writing.
Fev 2024 - Mar 2024
- Description: prediction of credit defaulting using data from credit contracts and bureaus.
Jan 2024
- Description: Identification of digits from a dataset of tens of thousands of handwritten examples.