Alejandro M. Simón Sánchez

Experience

Aug 2020 – Present | UCLM, Albacete

Deep Learning & Time-Series: Research and development of advanced architectures for large-scale crop classification and land-use monitoring using Sentinel-2 time-series.
End-to-End Data Pipelines: Design of full-stack pipelines, from satellite image preprocessing to cloud-based model inference.
Software & GIS Integration: Development of scalable solutions, including a native QGIS plugin for automated field-level prediction.
Researcher in high-impact EU-funded projects: REXUS (H2020) and NEXUS-LABS (PRIMA).

PythonTensorFlowGoogle Earth EngineQGIS Plugin

Jan 2020 – Aug 2020 | YOTTA Desarrollos Tecnológicos

Angular / IonicNode.jsMongoDB

Jan 2019 – Mar 2020 | UCLM (Institute for Regional Development)

European project H2020 (BIOPLAT-EU): Design and implementation of online WebGIS tools.
Requirements & Analysis: Requirements analysis and technical documentation.
International Collaboration: Communications and meetings with global partners (FAO, WIP, Geonardo, Joanneum Research, etc.).

Crop Classification & Irrigation Mapping (Mancha Oriental & Italy): Deep learning models using Sentinel-2 image time series for agricultural monitoring and irrigation assessment across study regions.
eo-inference-gateway & qgis-client: Service built with FastAPI and Docker for Deep Learning inference in Earth Observation (Sentinel-2), featuring spatial time-series extraction. Paired with a custom QGIS plugin for backend interaction. (Open Source release pending).

CompanySpider Architecture: Autonomous scraper acting as a Google Maps agent combined with web scraping (Python, Selenium, BeautifulSoup). Supported by a MongoDB backend and an Angular/OpenLayers admin dashboard.
Multi-level Routing & Indoor Geolocation: Proof of Concept for indoor geolocation utilizing Bluetooth beacons. Web-based topology editor (Angular, OpenLayers) for multi-level buildings featuring Dijkstra-based pathfinding with condition-based filtering.
Dynamic Public Data Extraction: Automated extraction pipeline for 24h pharmacy data, bypassing dynamic municipal dropdown forms lacking public APIs.

Convolutional Neural Networks for Agricultural Land Use Classification from Sentinel-2 Image Time Series
Remote Sensing • Oct 2022
Conference proceedings and papers on deep learning applied to agronomy (pending update & submission)...

Current | UCLM

Funding & Award: UCLM Own Research Plan Predoctoral Fellowship.

2019 – 2020 | UCLM

Grade: 9.66/10. Extraordinary End-of-Studies Award (Feb 2021). Final project: "System for automatic collection of business data from the web".

2014 – 2019 | UCLM

Grade: 8.32/10. Scholarship of Excellence (Dec 2019). Mention in computation / data science.

Professional Certifications & Digital Competences

Applications of Remote Sensing and GIS in Agriculture (9 ECTS) – (UCLM, 2019).
DIGCOMP 2.2 Level C2 Digital Competences Accreditation (UCLM, 2026).
ArcGIS Online Introduction (30h) – (UCLM, 2026).
TensorFlow Developer Program – DeepLearning.AI.
Deep Learning Specialization (5 courses) – DeepLearning.AI.
Machine Learning (60h) – Stanford Online (Coursera, 2020).
Scientific Programming with Python (20h) & MATLAB Level I – UCLM.

Hardware, Low-Level & Open Source Projects (Click to expand)

udpbd-server cross-compilation: Ported and cross-compiled for ARM7 32-bit architecture to integrate into Luckfox Pico. [GitHub ↗]
PsNee Fork: Fixed a hardware race condition by controlling the console's reset signal to ensure proper Arduino initialization, retaining the bootloader for easy flashing. [GitHub ↗]
Capcom Home Arcade Mod: Hardware reverse-engineering guide to add a physical USB port to use the arcade stick as a standard gamepad. [Guide PDF ↗]
Atari 2600 S-Video / Composite Mod Design: Custom circuit schematics designed for Atari 2600 (PAL/NTSC compatibility), integrating a 2N3904 transistor stage, passive mixing networks, and an FMS6400 filter/driver IC with proper 75Ω impedance matching. [Schematic PNG ↗]
SNEdge Collaboration (Thunder Technologies): Hardware contribution to improve video output on Super Nintendo/Famicom. Reverse-engineered undocumented board traces using a multimeter (Issue #6 ↗) and proposed low-capacitance shunt capacitors to eliminate board noise without disrupting mod functionality (Issue #7 ↗]).
N64RGBVDC Noise Troubleshooting (cosicasF9): Collaborated on resolving a voltage drop / noise issue on Nintendo 64 RGB mods by isolating the noisy 3.3V power rail and implementing a dedicated linear regulator (AMS1117) powered from the 5V line. [Video Ref ↗]

Community, Volunteering & Languages

Community & Volunteering: Co-Founder & Co-Organizer of RetroAlba (2013-2016). Tutor of an intern in the Dual Formation project for improving the employability of people with autism (Autismo España).
Languages: Spanish (Native), English (Full Professional Proficiency / B2 Certified 2012, daily use in EU projects & research), French (Basic), Chinese (HSK2).