54 skills · backed by entries on this site
Skills
What I can do, grouped by area. Each skill links to the work where I used it, so the claim comes with its evidence.
used in academic work used in industry
These skills are distilled from the work listed on this site. Each one links to the activities where I actually used it, so you can check the context rather than take a keyword's word for it. The year span next to a skill is computed from those activities.
Verification & safety assurance
How I make sure a system does what it should, especially when it is autonomous and safety-critical.
Scenario-based testing
System-level test scenarios: initial scenes, trajectories, oracles.
2023 – now
Model-based testing
Deriving test suites from formal or graph-based models of the domain.
2023 – now
Search-based testing
Casting test generation as optimisation (NSGA-II/III).
2024 – 2025
Simulation-based testing
Executing generated scenarios in maritime and robotics simulators.
2024 – now
Safety assurance
Building evidence that a system complies with rules such as COLREGS.
2024 – now
Runtime monitoring
Monitors that observe a running system and flag rule violations.
2022 – now
Coverage & completeness metrics
Measuring whether a test suite covers every semantically distinct case.
2024 – now
Temporal logic & automata
Parametric timed automata, metric temporal graph logic, complex event processing.
2022 – 2024
Design-space exploration
Heuristic exploration of model spaces to find valuable test sequences.
2022 – 2024
Constraint satisfaction
Geometric and logical constraint problems solved by search and sampling.
2024 – now
Verification & validation
Test protocols and validation for production software and embedded systems.
2022 – now
Model-driven engineering
Modelling languages and tools I use to describe systems precisely enough to reason about them.
Metamodelling
Domain metamodels for maritime actors, relations and COLREGS situations.
2024 – now
Graph models & constraints
First-order graph constraints and graph pattern matching.
2022 – now
Refinery
Graph solver for generating consistent models; I contributed monitors and use it daily.
2022 – now
VIATRA
Incremental graph queries and model transformations.
Xtext / Xtend
Textual DSL development on the Eclipse stack.
Autonomous & cyber-physical systems
Systems that sense, decide and move, whether at sea, on the road, or on a lab floor.
Maritime autonomy & COLREGS
Encounter geometry, relative bearings, give-way/stand-on roles, MASS code.
2024 – now
Motion & trajectory planning
RRT*-based planning of COLREGS-compliant target-ship trajectories; vehicle manoeuvres.
2022 – now
Gazebo & ArduPilot SITL
Software-in-the-loop vessel simulation.
2025 – now
MQTT & MAVLink
Messaging used to drive simulated and real vessels.
2025 – now
IoT telemetry
Compact message serialisation, performance monitoring and reporting to a server.
2021 – 2025
Software engineering
Building, shipping and maintaining software that other people depend on.
Python
My main language for research prototypes and backends.
2018 – now
Docker
Containerised multi-service stacks (Compose, spawned worker containers).
2022 – now
CI/CD
Pipelines from commit to deployment and monitoring (Jenkins, Bitbucket).
2022 – now
Software architecture
Modular, service-based tool architectures; mechanical/software co-design.
2021 – now
Requirements engineering
Eliciting and analysing requirements with international stakeholders.
2022 – now
SQL databases
PostgreSQL, MySQL, SQLite, MS SQL Server, InfluxDB.
2018 – 2024
Data & AI
Learning-based methods and data tools I have applied in projects.
Deep learning
CNN-based imitation learning; transformer models for NLP.
2021 – 2024
NLP & named-entity recognition
Information extraction from Hungarian legal text.
2018 – 2022
Evolutionary algorithms
NSGA-II / NSGA-III for multi-objective constraint solving.
2024 – now
Probabilistic sampling
Rejection sampling with the Scenic framework, sampling-region construction.
2024 – now
Topological data analysis
Persistence diagrams and clustering.
Human-centred AI
Explainability, fairness and dataset bias (HCAIM programme).
2022 – 2024
Graph & semantic data
Neo4j, RDF, OWL.
Research, teaching & leadership
The work around the work: evaluating, writing, presenting, teaching and leading.
Empirical evaluation
Designing experiments on completeness, diversity, scalability and speed.
2024 – now
Scientific writing
2022 – now
Conference presentations
2022 – 2026
Teaching
2024 – now
Thesis supervision
2024 – now
Project management
2022 – now
Agile / Scrum
2025 – now
Languages
Hungarian
Native
English
Proficient
Swedish
Basic