Skip to content
DSML.KZ
AS

Arystan Shokan

@arctanConnected with DSML. This person's Telegram account has been confirmed through DSML Kazakhstan.

Mathematician and Deep Learning Engineer

Geometric Deep Learning and Topological Data Analysis. Persistent homology, Laplacian are my favorite algorithms, and Manifold learning is my favorite type of learning. I also like graph theory and Graph Neural Networks

Open to

Learning with othersHelping with community projectsHelping organize meetupsResearch collaborationPaper discussions

Featured

Career

Experience: 4 years

Experience

Operation Research and ML Engineer

Navix

Mar 2026 — Oct 2026

Using optimization algorithms, like Linear Programming, Dynamic Programming, ALNS, ACO, Greedy Methods to optimize problem similar to TSP and VRP. Responsible for formulation of a Loss function, linearization, fine tuning all hyperparameters

Software Engineer

Bereke Bank

Feb 2024 — Dec 2024

• Developed and launched the B-Business mobile App as a key member of the core platform team • Integrated functional programming methodologies, which reduced runtime errors by 40% and improved overall developer satisfaction • Presented on the mathematical foundations of functional programming at local tech meet-ups • Spearheaded code reviews and architectural committee initiatives, successfully reducing legacy code by 70%

Research and Development Engineer

Innoforce Systems

Oct 2022 — Dec 2023

• Researched different methods of accessing GPS on Android devices that resulted in more stable access of GPS data using sensors and location providers in the Android OS, achieving accuracy up to 5 meters • Wrote multilateration algorithm via WiFi-RTT technology to receive geo position • Algorithms were used for post terminals in buses with Android 7 operating system • Implemented Kalman Filter and Prediction-Correction algorithms

Software Engineer

Kazdream Technologies

Nov 2020 — Sep 2022

• Reduced render time from 3 seconds to 3 milliseconds using reactive programming and caching • Wrote clustering algorithm that resulted in the rendering of 10x fewer UI elements using concepts of Graph Theory • Implemented map module applying concepts of Differential and Analytical Geometry • Added functionality to draw customizable markers on a map using functional programming principles

Education

MA Mathematics

Boston University

2025 — 2026

Specialization in Geometry: Manifolds, Sheaves, Varieties, Graphs. Thesis in Combinatorial Hodge Theory about graphs as a special case of simplicial complexes and analysis of chow ring of matroids

BSc Software Engineering

Astana IT University

2020 — 2023

Thesis in combinatorial optimization in graph theory using Linear Programming, Genetic Algorithms and Graph Neural Networks

Skills

PythonMLGeometryMathematicsJulia