Machine learning & AI
Neural networks, deep learning, GANs, LLMs and AI agents, with statistics and probability as the foundation.
I work where machine learning meets real-world engineering: from biometric recognition research to LLM-powered agents and the backend systems that run them.
I'm deeply passionate about maths and computer science, and I enjoy solving hard problems in collaborative environments. I've been around computers since I was a kid, and what still drives me is using them to solve real problems that improve people's lives.
I spent years doing research on fingerprint, palmprint and speaker recognition: image processing, statistical modelling and deep learning applied to forensic biometrics. Today I build backend systems and AI tooling, and I advise teams on LLM agents, cloud architecture and delivery pipelines.
When I'm not researching or building, I'm working on open source, or playing blitz chess.
{ "name": "Manuel Aguado Martínez", "role": "Computer Science Engineer", "education": "BSc. CS, University of Havana", "focus": [ "machine learning", "LLM agents", "biometrics", "backend systems" ], "stack": ["python", "c/c++", "c#", "js"], "speaks": ["es", "en"], "chess_blitz_fide": 1932 }
Neural networks, deep learning, GANs, LLMs and AI agents, with statistics and probability as the foundation.
Fingerprint and palmprint recognition, minutiae extraction, speaker recognition, and evaluation of biometric systems.
Enhancement and contextual filtering, frequency-domain analysis, segmentation, thresholding and texture analysis.
Python, Django, Flask, REST, microservices, Docker, AWS serverless and CI/CD pipelines.
A corporate AI platform that turns scattered documents, data and processes into operational knowledge and traceable decisions.
Evaluates biometric and binary classification systems, reporting EER, ROC/DET and CMC curves the way they should be reported.
NumPy/SciPy implementations of state-of-the-art image thresholding algorithms.
→ more on researchgate
Want to work together, talk research, or play a game of blitz?
$ ./contact --form contact form: coming soon™ $ ls ~/links github linkedin researchgate $