Tabish A.

Applied mathematician
and ML researcher

I work on complex systems, forecasting, and computation—connecting mathematical structure with software and real-world modelling.

xy∂u/∂t = ν ∇²uu(x,0) = u₀(x)‖∇f(xₖ)‖ → 0L(θ)t∇²uλₙ∫ f dxθ ← θ − α∇L

Based in Hamburg / Europe

Applied Mathematics + ML

Research, teaching, writing

Open to collaboration

Selected Work

See all →

LLM Systems

Prompt Optimization Framework at Suncorp

Built a practical framework to improve prompt quality, evaluation, and reliability.

See project

Research

KAN / Forecasting Research

Explored interpretable neural structures for forecasting complex dynamics.

See project

Applied Math

Mathematical Modelling Project

Designed and tested a modelling pipeline for hard-to-predict real-world systems.

See project

Selected Writing

Read more →

On Uncertainty, Forecasting, and Structure

A reflection on where mathematical assumptions meet real-world behavior.

Read post

Notes on Building With Models

Practical thoughts on turning theory into tools that people can actually use.

Read post

Personal Note

I am interested in systems that are hard to predict, whether they arise in mathematics, machine learning, or human life.

Let’s collaborate

Open to research, and writing collaborations.