AI & Data Solutions Engineer

Neil
Basson

Industrial engineer by training. AI solutions practitioner by focus.
Building production-ready systems that drive real business outcomes.

View Work Get in Touch
Scroll

About

Systems thinker.
AI builder.

Currently embedded at group level within a multinational manufacturing group, building data infrastructure and AI capabilities that span multiple business entities.

I work across the full stack — from lakehouse architecture and ML pipelines to generative AI systems and agentic workflows — always with a focus on solutions that are production-ready, explainable to stakeholders, and tied to real business outcomes.

🇳🇱 🇿🇦 Dutch & South African 📍 Amsterdam, Netherlands

Education

University of Amsterdam

MSc Business Information Technology Management

University of Amsterdam · 2025 – 2027

University of Amsterdam

MSc Data Science & Business Analytics

University of Amsterdam · 2024 – 2025

Stellenbosch University

BEng Industrial Engineering

Stellenbosch University · 2020 – 2023

Capabilities

Built for the
full AI stack.

⚡

Core Stack

PythonSQLSparkGitMicrosoft FabricAzure
🧠

AI & Machine Learning

PyTorchTensorFlowKerasScikit-learnClassical MLTime Series
🤖

Generative AI & Agents

LLMsAzure OpenAIClaudeRAGMulti-Agent SystemsPrompt EngineeringAzure AI Search
🏗️

AI Architecture & Delivery

End-to-End AI DesignMLOpsAPI IntegrationModel DeploymentSecure Enterprise AI

Experience

Where I've
delivered.

Data & Machine Learning Engineer

TenCate Grass

Sep 2025 – Present
  • Built a Microsoft Fabric lakehouse (bronze → silver → gold) with automated ingestion pipelines using Data Factory, PySpark & Delta Lake.
  • Developed a customer segmentation model with XGBoost & Random Forest on geographic and spend features to surface high-value customer groups and seasonal patterns.
  • Engineered a time series sensor data workflow for a high-volume manufacturing line, integrating PLC streams with window-based statistical thresholds linked to quality defects and yield performance.
  • Mapped cross-company process flows in SQL and Python to surface data gaps across production, logistics, and sales entities.

Lean Manufacturing Data Analyst

Heineken Beverages South Africa · Internship

Mar – Nov 2023
  • Analysed production line data with Python & R to identify bottlenecks in Heineken's box wine packaging line using statistical modelling.
  • Quantified waste reduction opportunities, proposing targeted improvements projected to reduce production waste by ~2%.
  • Delivered findings via interactive Power BI dashboards enabling non-technical stakeholders to act on operational trends.

Junior Operations Data Scientist

Ocean Guard Africa

Jun – Dec 2022
  • Built a regression model with Scikit-learn to predict equipment failures, reducing unexpected downtime by 15% through targeted maintenance planning.
  • Automated repetitive data processing with Python & pandas, cutting manual reporting time by ~6 hours per week.
  • Integrated Excel procurement data via openpyxl, contributing to a 20% reduction in procurement cycle times.

Projects

Selected
work.

MSc Thesis (Management) · Current

AI Governance & Deployment Readiness Framework for Clinical ML

Developing a governance and deployment readiness framework for operationalising a lung cancer classification model in a clinical radiology setting. Maps the ML pipeline against EU AI Act requirements, ISO/IEC 42001 controls, and FDA/IMDRF medical AI lifecycle guidance.

EU AI ActISO/IEC 42001Design Science ResearchMLOpsResponsible AI
MSc Thesis (Data Science) · 2025

Lung Cancer Segmentation & Classification Pipeline

End-to-end segmentation-classification pipeline in PyTorch processing CT scans to predict lung tumour malignancy. Multi-branch deep neural architecture with Grad-CAM interpretability on the 130 GB LIDC-IDRI dataset.

PyTorchCNNGrad-CAMDICOMResponsible AI
2024

Brain Tumour Classification & Segmentation

Custom deep learning pipeline in TensorFlow/Keras for simultaneous MRI segmentation and classification, combining CNNs with data augmentation, structured label extraction, and SGD variant experimentation.

TensorFlowKerasCNNMRISegmentation
2024

Predictive Modelling & Feature Engineering

Built and compared regression models (Ridge, Lasso, KNN, Kernel Ridge) with PCA dimensionality reduction and GridSearchCV hyperparameter tuning, evaluated with structured cross-validation for generalisability.

Scikit-learnPCAGridSearchCVCross-validation

Contact

Let's build
something.

Open to collaboration, consulting, and interesting problems.
Reach out and let's talk.

[email protected] LinkedIn ↗