Hernan Cordova

I build graph-based estimators, adaptive message-passing algorithms, and reference implementations for difficult inference problems where standard solvers are biased or fail to converge. At Eindhoven University of Technology, I design curvature-adaptive min-sum estimators and lead Ministry of Defence-funded optimisation research for logistics and resource allocation.

At BrainBox Digital and MGI Phikari Group, I led applied machine-learning and optimisation programmes across healthcare, industry, energy, and finance. My work on pharmaceutical distribution reduced backlog by 22% and cost by approximately 18% within nine months, supported by the data pipeline and data-quality framework behind it.

I bring a PhD in estimation, optimisation, and statistical signal processing, plus earlier hands-on field engineering experience building GPS radio-frequency systems and national-scale fibre-optic infrastructure. I build algorithms in Python or MATLAB, benchmark them rigorously, and use deterministic and learned methods according to what performs on the failure cases that matter.