A selection of projects where parallel computing, numerical methods and artificial intelligence become working software systems.
Each project focuses on the problem, my contribution, the engineering decisions behind the implementation and the results obtained.
High-Performance Computing · GPU Computing · Performance Engineering
A performance-oriented project focused on transforming a compute-intensive workload into parallel CPU and GPU implementations. The project explores workload decomposition, memory access, synchronisation and execution configuration through comparative benchmarking.
Focus: correctness validation, execution-time analysis and comparison between sequential and parallel implementations.
Scientific Computing · Finite Elements · Distributed Computing
A distributed finite-element solver for two-dimensional fluid–structure interaction, coupling incompressible Stokes flow with linear elasticity.
The project covers mesh generation, finite-element discretisation, distributed matrix assembly, iterative solvers, numerical validation and parallel performance analysis.
Result: validated across multiple MPI ranks and evaluated through a strong-scaling study on a problem with approximately 195,000 degrees of freedom.
Numerical Linear Algebra · Scientific Computing · Numerical Analysis
A collection of computational experiments exploring matrix algorithms, numerical stability, convergence behaviour and iterative methods for solving linear algebra problems.
Focus: connecting mathematical formulations with practical implementation, numerical accuracy and computational cost.
Applied AI · Time-Series Forecasting · Deep Learning
A project-based exploration of the complete forecasting pipeline, from raw temporal data and statistical baselines to neural-network training and model-performance assessment.
The work covered temporal windowing, chronological dataset splitting, autoregressive benchmarks, feed-forward and recurrent neural networks, GPU training, hyperparameter tuning and multi-step forecasting.
Embedded Systems · IoT · Distributed Software
A series of embedded projects involving reactive systems, cooperative scheduling, sensors, actuators and distributed device-to-service communication.
The work includes an interactive memory game, a task-based smart bridge controller and a distributed river-monitoring system using connected embedded devices.
Applied AI · Python · Software Integration
Experimental work on Python-based AI and automation systems, including conversational workflows, external-service integration and the evaluation of different large-language-model providers.
The work explored practical concerns such as embeddings, function calling, streamed responses and integration with existing software components.
A full-stack academic web application exploring user management, social interactions, content publication and persistent data.
Database modelling and implementation work covering conceptual design, relational translation, SQL queries and data-integrity constraints.
A team-based software project focused on object-oriented modelling, modular architecture, separation of responsibilities and collaborative development.