Antonio Reche

Hi! This is Antonio Reche. I like to train deep neural networks and use them to solve scientific problems. Here you can find my projects, experiences, and other things about me. You can also read my blog where I'll be documenting my thoughts, learning, experiences, and any other random stuff I do.

Photo of Antonio Reche

Research

  • Thumbnail for Polar faculae and their relationship to the solar cycle

    Polar faculae and their relationship to the solar cycle

    A. Reche, A. Pastor Yabar, A. B. Griñón-Marín

    Astronomy & Astrophysics, 2025

  • Thumbnail for Solar filament detection, classification, and tracking with deep learning

    Solar filament detection, classification, and tracking with deep learning

    A. Reche, C. Cid

    SPAICE 2024 (Conference paper), 2024

Experience & Education

University of Alcalá, Nov. 2023 - Present

Research Fellow

  • Since Dec. 2024 I'm doing a PhD in Science. I'm using deep learning to detect and analyze solar filaments over time with the objective of forecasting space weather phenomena.
  • After Medicsen, I joined the Space Weather Group to work on computer vision for detecting solar filaments.
  • Calar Alto Observatory, Nov. 2022

    Observational Internship

    Conducted a remote observational campaign to study the orbital properties of the binary star system V1301 Cas.

    Valencian International University, 2022 - 2023

    Master's Degree in Astronomy and Astrophysics

    While working at Medicsen, I enrolled in a second Msc 🤦🏼. I ended up doing research in solar physics using deep learning to detect and analyze magnetic structures in the Sun's poles.

    Medicsen, 2017 - 2022

    Chief Analytics Officer (CAO)

    I led the development of all the AI and data analysis technologies at Medicsen, a medical startup developing a needle-free drug delivery device. I spent most of my time building SugginAI, an AI-powered glucose prediction system that got certified as a medical device by the Spanish regulatory organism AEMPS. It was used by real patients and got coverage in some news: ConSalud, Nubentos.
    I also researched, built, and got to explore some other ideas: a voice interface where patients could just talk about their meals, drug intake, and physical exercise instead of manually logging everything, simulations of the human metabolism for drug delivery optimization... I also initiated the development of a closed-loop artificial pancreas for automated insulin delivery.

    Almería City Council, May - Aug. 2017

    Database Developer

    Started the development of a database and management software for the city's animal center.

    University of Granada, 2015 - 2016

    Master's Degree in Physics and Mathematics

    I developed a neural network 'unlearning' process to simulate the emerging phenomenology of Alzheimer's disease. The thesis was called 'Emerging Phenomenology in a Physical-Mathematical Model of a Degenerative Neural Network'.

    JEDA Granada, 2013 - 2016

    Science Communication and Outreach

    JEDA Granada was a junior enterprise dedicated to scientific outreach. During my undergrad and first Msc. I was part of it. I conducted astronomy workshops, observations, and taught some software courses.

    Università Degli Studi di Torino, 2013 - 2014

    Bachelor's Degree in Physics: Erasmus Programme

    I did one year abroad in Italy while studying my BSc in Physics. This is where I first discovered deep learning, through an elective course I took called 'Reti Neurale' (Neural Networks).

    University of Granada, 2010 - 2015

    Bachelor's Degree in Physics

    Thesis: Dynamic Light Scattering and Brownian Motion for Measuring the Size of Colloidal Particles.

    Side Projects

    Conferences & Seminars