I am dipping my toes in PDES (Parallel Discrete Event Simulation) and neuromorphic computing under the supervision of Professor Christopher D. Carothers. I am building a spiking neural network (SNN) simulator for HPC (High Performance Computing) called doryta.
In my masters at UNAL, I studied in depth how to check for errors produced by mismatched tensors in machine learning applications. The result of my work, with the help of my advisors Fabio González and Felipe Restrepo Calle, was an abstract interpreter which could be used as a linter to check for common errors in tensor operations using NumPy in Python. Both, the code (interpreter) and the thesis document, can be found online in my github page.
I also worked for over two years as developer/researcher for a chemistry and maths collaboration project at UNAL, Germina under the lead of professor Jesús Ágreda and the maths expertise of professor Juan Andrés Montoya. We explored different algorithms that might help on finding chemical networks that explain homochirality, a phenomenon found in nature and argued to be a key part of life itself. The result of the colaboration can be found at Gitlab Homochirality Group.