Pretoria, South Africa

Hi! I amArné Schreuder.

AI researcher and entrepreneur. I run Neurulation, an AI lab researching machine learning neuroscience, and serve as CTO of Taxnetix.
I am a PhD student and former lecturer in computer science at the University of Pretoria.

Founder & CEO - Neurulation
CTO - Taxnetix
PhD Student - University of Pretoria

Download CV Contact

About

I have spent the last 15+ years building software professionally, most of which focused on machine learning - production ML platforms for enterprises such as MTN and EPI-USE, real-time AI video processing, and agentic AI systems.

My current work centres on one question: can artificial neural networks support and repair biological ones? At Neurulation this takes the form of research into auxiliary neural networks for neuron function replacement, with Alzheimer's disease as the long-term target. My PhD research on biologically plausible artificial neural networks provides the scientific foundation.

At Taxnetix I lead technology strategy and the development of IntelliTax: an agentic-first, end-to-end tax intelligence platform for individuals and organisations in South Africa. I lectured software engineering and AI at the University of Pretoria for five years, and supervised more than twenty postgraduate research projects.

15+: years in software
10+: years in ML/AI
20+: researchers supervised
3: peer-reviewed publications

A new kind of software.

Neurulation is an AI lab based in South Africa. We do research in machine learning neuroscience, working toward solving Alzheimer's disease by engineering auxiliary neural networks capable of neuron function replacement. We also specialise in agentic AI for recursive self-improvement and machine learning science.

Machine learning neuroscience: Research at the boundary between artificial and biological neural networks - how learning systems can model, mimic and support the brain.
Neuron function replacement: Auxiliary neural networks designed to take over the function of failing neurons, aimed at Alzheimer's disease.
Agentic AI & recursive self-improvement: Agentic systems that design, evaluate and refine AI - accelerating the pace of machine learning research.
Machine learning science: Publication-grade research methodology: hyper-heuristics, spiking neural networks and optimisation.

Experience

Present

Founder & CEO - Neurulation

AI lab researching machine learning neuroscience - auxiliary neural networks for neuron function replacement - and agentic AI for recursive self-improvement.
Since Jan 2026

Chief Technology Officer - Taxnetix

Technology strategy, AI innovation and product architecture. Lead of IntelliTax: an agentic-first, end-to-end tax intelligence platform for individuals and organisations in South Africa.
2020 - 2025

Client Engagement Lead · Senior Software Engineer AI/ML - EPI-USE Labs/Advance

Founded System42, the group-wide enterprise agentic-AI platform. Led MLOps for MTN's Advanced Analytics platform on Azure Databricks; real-time AI video processing (super resolution, frame doubling, scintillation removal); technical lead of the Innovation Africa @UP Information Hub, serving 3000+ users.
2020 - 2025

Lecturer, Software Engineering & AI - University of Pretoria

Lectured COS 301 and COS 731 (Software Engineering, undergraduate and honours). Supervised 20+ postgraduate research projects, largely on spiking neural networks and biologically plausible machine learning.
2016 - 2019

Chief Executive Officer - Compiax

AI product company: Artofact.io (AI-optimised business management), PropertySync.io, and NLP platforms for the customer-experience industry.
2013 - 2019

Senior Software Engineer - Consulta

Architected Analytiax and Datiax - micro-service NLP platforms for text and sentiment analysis powering customer-satisfaction analytics, including the South African Customer Satisfaction Index.
2011 - 2013

Chief Executive Officer - Arn3

First venture: full-stack software engineering and consulting.

Research

PhD student in computer science at the University of Pretoria, supervised by Prof. Anna Bosman, specialising in biologically plausible artificial neural networks. MSc in computer science/AI (cum laude), BSc Hons (cum laude).

Publications

Training feedforward neural networks with Bayesian hyper-heuristics
Information Sciences, Vol. 686, 2025 - with A.S. Bosman, A.P. Engelbrecht, C.W. Cleghorn
Training Graph Neural Networks with Particle Swarm Optimisation
SACAIR, 2023 - with R. Shen, A.S. Bosman, M. Krzywda, S. Łukasik
Customer Centric Artificial Intelligence - Text & Sentiment Analysis with Deep Neural Network Learning
South African Journal of Marketing Research, No. 13, 2017

Research groups

Computational Intelligence Research Group (CIRG), University of Pretoria
Innovation Africa @UP
Forestry & Agriculture Biotechnology Institute (FABI)

Supervision

20+ MSc and Honours research projects - spiking neural networks, neural numerosity, GNN optimisation, vision transformers, few-shot learning.

Speaking & Awards

Deep Learning Indaba · 2024

“An Introduction to Spiking Neural Networks and Biologically Plausible Machine Learning”
Chief Data Officer & Information Quality Symposium · 2024

“The Impact of AI on Data Quality for Large-Scale ERP System Migrations”