Projects

How do you turn hours of manual peak picking into an analysis that anyone can rerun and audit? This case study presents a Python pipeline that reads raw GC-MS acquisitions, corrects the baseline, detects and integrates peaks following the vendor methodology, and delivers interactive reports and tidy tables, validated against results produced by hand.

Can a genuine computational fluid dynamics solver run at interactive frame rates inside a browser tab? Web-LBM is a 2D wind tunnel built on the Lattice Boltzmann Method: a C engine compiled to WebAssembly, a zero-copy WebGL2 rendering pipeline, a turbulence model, and multithreading, all served from a 100 % static page.

How do you turn a textbook flocking model into a simulation that scales to tens of thousands of agents and renders like a fluid? This case study walks through the boids algorithm, four progressively faster solver backends benchmarked side by side (up to about 330 times faster than naïve Python), and a GPU visualization engine built with ModernGL.

How do you design an airfoil that performs well throughout an entire flight? This case study showcases a Python toolkit that explores thousands of candidate geometries and leverages evolutionary optimization to uncover the most effective aerodynamic trade-offs between lift and drag.

A full-stack scientific web app that derives the dimensionless groups governing a physical problem with the Vaschy–Buckingham (Π) theorem. The real Python + SymPy engine runs entirely in the browser via WebAssembly.