Nadav Port

Backend software engineer with a healthy interest in finance

Welcome to my personal website, where I share some of my past work and academic projects
I am a software engineer with experience in server-side Python and DevOps. I recently graduated from Imperial College London with a Master’s in Applied Mathematics, where I enjoyed diving deep into mathematics and machine learning.
I have a strong interest in software, finance, ML, and math, and especially at their intersections. Currently, I am looking for job opportunities in London within these fields, so please feel free to contact me if you have any matching opportunities!

Essays

When AI is not good enough

Past Projects

Masters Thesis - Neural Operators

Numerical analysis of the theoretical neural operator architecture for learning PDEs, implemented using PyTorch.

Fluid simulation using Python

Simulating complex function contours to create 2D fluid flow simulations using Sympy and Manim!

Numerical Analysis of Biological PDEs

Conducting both numerical and mathematical analyses, I explored a model depicting a biological process with unconventional boundary conditions. Employing finite difference methods, I crafted a suitable numerical scheme and implemented it using Numpy.

Mathematical Modeling of Battery Phase Transition

Collaborating with my advisor, we developed a mathematical model capturing the unique phenomenology of phase transition in a nickel-hydroxide electrode within a battery. I conducted in-depth analyses using Matlab and Python.

Deep Learning Feed-Forward Network

Implemented a feed-forward deep learning network strictly in NumPy, following backpropagation straight from the math. I initiated, trained, and analyzed the network's performance, specifically its ability to recognize MNIST digits and polygon edges.

Analysis of Chaos in Chemical Reactions

For the final project in a Nonlinear Dynamics module, I analyzed a chaotic system of equations based on the Willamoski-Rossler equations. Exploring numerous parameters, I selected a few and detailed their influence on the system.

Setting Up Real-Time Data Science Workflow Tool

During my army service, I focused on real-time, continuous data processing. In this project, I configured and tested the 'Argo Workflows' workflow/pipeline programming tool on our Kubernetes cluster. I found the software valuable and documented its basic usage.

Generic Computer Networking Security Layer in Openshift

Developed a generic networking security layer utilizing NGINX and the OIDC protocol for applications running on OpenShift. The objective was to provide OIDC protection without developers needing to implement it directly into their code.