Hello, I'm Volodymyr Halian
Building intelligent
systems with
machine learning.
AI & Machine Learning Developer exploring deep learning, neural networks, and the ideas behind intelligent systems. From mathematical foundations to real-world applications.
AI & Machine Learning Developer
Ideas, explored.
Models, understood.
Exploring the mechanics of learning through academic work, code, and careful experimentation.
All ML projectsA toolkit built on understanding.
Practical tools. Mathematical foundations.
A habit of asking why.
Machine learning
COURSEWORK & EXPLORATION- Classification & regression
- Clustering
- Dimensionality reduction
- Feature engineering
- Model evaluation
Deep learning
COURSEWORK & PRACTICE- PyTorch
- FNN & CNN architectures
- Backpropagation
- Gradient descent
- Neural network experiments
Programming & data
PRACTICAL TOOLKIT- Python · NumPy · Pandas
- Matplotlib · scikit-learn
- JavaScript · TypeScript
- React · React Native
- Git & software prototyping
Mathematical foundations
ACADEMIC FOUNDATIONS- Linear algebra & calculus
- Probability & statistics
- Numerical optimization
- Bayesian inference
- Algorithms
It's all about
paying attention.
Causal vs Full Attention
The same sequence.
Two ways of seeing it.
How does access to context change what a transformer can learn? My planned academic work explores attention masking, token dependencies, and information flow.
Causal attentionLearning from the past, one token at a time.
Full attentionUnderstanding each token in its complete context.
Try the matrix → Switch modes and select a query token.
Can attend to The · model · learns. Future positions are masked.
Binary access mask · explanatory visualization, not learned weights
Predicting in representation space.
JEPA, world models, and the dynamics of gridworlds.
Explore the ideaSmall experiments. Deeper understanding.
01Probability & Bayesian inference
Priors, likelihoods, posterior distributions, and reasoning under uncertainty.
Bayesian inference · Statistical samplingAcademic and technical exploration. Individual implementations and results can be added as documented project artifacts.
02Optimization & numerical methods
Gradients, line searches, Armijo and Wolfe conditions, and constrained optimization.
Calculus · KKT · Numerical optimizationAcademic and technical exploration. Individual implementations and results can be added as documented project artifacts.
03Data & mathematical explorations
PCA / ICA, recommendation methods, preprocessing, and careful model evaluation.
Linear algebra · Data analysisAcademic and technical exploration. Individual implementations and results can be added as documented project artifacts.
04Algorithms & reliable software
State machines, algorithmic reasoning, testing, and making computational behavior explicit.
Algorithms · Software testingAcademic and technical exploration. Individual implementations and results can be added as documented project artifacts.
Useful ideas belong
in the real world.
Product thinking, software, and the practical constraints that make engineering interesting.
A more considered AI workflow.
Structured inquiries. Explicit business rules. Human review.
Making time easier to plan.
A calendar and task app explored with React Native and Expo.
Designing repeatable operations.
Wholesale catalogs, tasting coordination, and clearer handoffs.
Learning. Building. Connecting the two.
Johannes Kepler University Linz
Bachelor's Program in Artificial Intelligence
Linz, Austria
Machine learning, neural networks, probability, linear algebra, optimization, reinforcement learning, algorithms, and software engineering.
Upcoming seminar & practical workMousse Bakery
Digital projects, AI-assisted workflows & operations
Chicago metropolitan area, USA
Website-related work, AI-assisted customer communication, wholesale coordination, and practical product and workflow design.
Explore the engineering workCurious about the theory.
Driven to build with it.
I'm Volodymyr, an Artificial Intelligence student at JKU Linz. I enjoy understanding how intelligent systems learn from data, then turning that understanding into code.
My interests sit at the intersection of deep learning, transformer architectures, and representation learning. I like studying the mathematics behind a model, implementing its ideas, and asking what the behavior actually tells us.
Alongside my academic work, I've been involved in real-world digital and business projects. Working with practical requirements has made me value clear systems, thoughtful validation, and useful software.
More about meA resume for the opportunity.
ML engineering, AI engineering, data science, or software.
The same experience, with the relevant work up front.
Let's build something
intelligent.
I'm interested in machine learning, AI research, and software engineering opportunities. Let's connect about internships, working-student positions, or technical collaboration.
Linz, Austria · Open to opportunities across Europe