I currently work to design new features, and optimize, Google's search engine.

My research in the Computer Architecture and Parallel Systems Laboratory (CAPSL) focuses on thedevelopment of techniques for execution of scientific programs. It spans from source code optimization,to compiler technology, to supercomputer execution models. Several publications have been producedby me, including two papers on how to optimize scientific programs, a series of talks aboutsupercomputing to the French Commission of Atomic Energy and several contributions to parallel codingtechniques that will be used in Los Alamos National Laboratory.

Neutron Transport simulations are the core of many simulations that deal with nuclear interactions.Unfortunately, computationally speaking, these simulations are very expensive. I reviewed thealgorithms in use, and, while working in the HPC-5 group, I proposed 14 new techniques that can speedup the simulation process. As a result, I developed a simplified Neutron Transport Solver, NMNT, thatincludes these new techniques, and that can be used to do research on supercomputer performance.

One of the most difficult design aspects of automatic software test tools is the generation of the inputsto the software being tested. I worked in a group to develop and build a more efficient input generatorto test pieces of software. As an aside, I became very familiar with the software and the hardware ofCyclops 64, a many core chip from IBM that features 160 thread units in a low power chip.Tecap: Circuit Design and Software Engineer (2002-2005)During the first years of my career, I championed a series of projects for industrial automation. I led theinitiative of entire automation processes while having under my supervision a number of technicians.The most representative systems that I built include artificial intelligence software for motion analysis,automatic controllers for a variety of machines, and several systems for processing signals received fromindustrial sensors.Research Projects2009 · Compiler transformations for data locality. · Languages for parallel execution. · Parallel execution models.2008 · Supercomputing Performance Tuning. · Scientific Algorithm Design and Optimization for New Multicore Architectures.2007 · Optimization for Multi Core Architectures. · Execution Models for Multi Core Architectures2006 · Simulation and verification of the Cyclops 64 A Switch hardware. · Source code optimizations for superscalar architectures2005 · Fast Hardware Algorithms for High Performance Scientific Computing

my years as a recent graduate, I participated as a researcher in a countrywide initiative todevelop an automated system for geographical verification of Colombia's Power Infrastructure. Theproject included the design and construction of an unmanned aerial vehicle equipped with multiplesensors that allowed remote monitoring of power lines.

Due to exceptional quality during my undergraduate years, I was hired as an Assistant Professor onlyweeks after graduation as an Electrical Engineer. I taught courses to hundreds of students in ComputerProgramming, Automatic Control Design, and Circuit Design.