A collection of articles and write-ups exploring engineering, architecture, and mathematical ideas.
A continuation of the previous article:
the formulas are mathematically correct, but real computation reveals something deeper — numerical stability.
Understanding this difference emerges not from the formula itself, but through reasoning and human intuition.
An article about using ChatGPT to restate an already-solved least-squares geometry problem in a more compact mathematical form. It connects the original engineering derivation with vectors, matrices, gradients, and the broader question of how AI can express known mathematics clearly without changing the underlying solution.
A real-world engineering story about using AI for refactoring, where a simple commit request turned into an unexpectedly competent next-step intervention. The piece connects that experience to Isaac Asimov's Foundation novels and the practical problem of building trust with advanced machines.
After 10 years exploring Rust architecture, I finally figured out how to build enterprise-grade applications that are truly Rust-native — thread-safe, scalable, and zero-runtime-cost.
Decade ago, I tried writing Rust like Java: mutexes, dynamic dispatch, atomic reference counting. The code ballooned, and I realized it wasn’t the Rust way, so I kept searching. Now, I’ve built a system that embraces Rust’s strengths: compile-time guarantees, smooth abstractions, and blazing performance.
If anyone asks whether Rust can handle complex enterprise applications, my answer is: yes. Source code is linked in the article.
Recently I’ve solved computer vision problem, and I’d like to share approach that I’ve invented to solve that problem. Typically such problem can be easily solved with use of machine deep learning and neuron networks. However I’ve got insight that this problem could be solved with non-trivial but very elegant solution, my knowledge in mathematics led me to use Least Square Optimization for Optical Flow representation of captured video stream, without need to train algorithm with real videos. This approach could automatically adopt to any environment as much as optical flow could provide accurate output.
Mathematical analysis about computation complexity – particularly comparing O(n log n) function with O(n) function, and some other theory about infinities.
Describes different branching strategies. Nowadays, Git-Flow is very popular and well-documented, however, there are a few other alternatives like GitHub-Flow and the Trunk-based approach exist and some teams tend to move away from git-flow, and this article will describe why some other branching models may work better for your and your team.
