Alan
Amaro
Physicist & AI Developer | R&D Specialist | Innovation Engineer
Translating complex science into deployable, high-impact AI products. From predictive molecular modeling to autonomous agentic systems.
About Me
I am a physicist and AI researcher focused on translating complex science into deployable, high-impact AI products. My expertise bridges full-stack technical execution with strategic product development, with a proven track record of building scalable systems for computational chemistry, materials science, and medical diagnostics.
My most significant achievements involve end-to-end ownership of scientific AI tools — from synthetic data generation and custom neural network architecture design to production-ready desktop applications and cloud-deployed agentic systems. I hold a Bachelor's in Physics from Universidad Nacional Autónoma de México (UNAM), where I specialized in Artificial Intelligence, Materials Science, Data Science, and Computational Chemistry.
I am driven by the challenge of turning groundbreaking research into reliable, user-centric technology. Actively seeking collaborations with innovators and organizations that want to leverage cutting-edge AI to solve the most pressing challenges in science and industry.
Projects
End-to-end AI solutions across computational chemistry, materials science, paleontology, and more.
AzoNet — Thin-Film UNET
Deep convolutional network predicting thin-film thickness from transmittance spectra with MAPE of 7% and R² of 0.93.
QSAR Predictive Modeling
Deep Neural Networks for predicting biological activity of Alzheimer's and antibiotic compounds. MAPE of 5.52%, R² of 0.81.
Nexus+ — B2B AI Assistants
Co-founded a SaaS startup deploying agentic AI assistants for the medical sector with RAG pipelines and cloud infrastructure.
Paleo-Net — 3D Fossil Reconstruction
3D Convolutional U-Net for restoring damaged fossil specimens. Presented at the XIX National Paleontology Congress.
CanSat Satellite Prototype
Led hardware & software development for a canned satellite. Real-time data visualization. 22nd out of 101 teams nationally.
XRD Crystal Classification
UNET for crystal system classification and lattice parameter prediction from XRD spectra. MAPE of 1%.
CICHAZ Research Internship
International collaboration on 3D tracking with OpenCV, Raspberry Pi networking, and synthetic simulation data.
Skills & Technologies
Programming Languages
AI & Machine Learning
Data & Infrastructure
Frameworks & Tools
Languages
Contact
Let's collaborate. Whether it's AI research, product development, or consulting — reach out.
Get in Touch
I'm actively seeking collaborations with innovators and organizations that want to leverage cutting-edge AI to solve pressing challenges in science and industry.