Hi, I'm Kristof Stroobants|

• Technical lead on GenLayer Studio, responsible for backend and SDK development in Python, blockchain integration and creating a seamless developer experience through CI/CD pipeline, testing, documentation, code reviews and architectural decisions
• Built the whole consensus mechanism for transaction processing (Finite State Machine)

• Building a platform that balances developer flexibility with protocol constraints
• Creating intuitive abstractions for complex blockchain concepts
• Ensuring seamless integration between frontend, backend and blockchain layers
• Maintaining comprehensive documentation for rapidly evolving features

This project makes my work publicly visible and demonstrates how I structure, document and maintain production-quality code. It shows my approach to developer experience and long-term maintainability in open-source environments.

Embedded software engineer responsible for verification and testing of safety-critical FPGA systems.

• Developed and simulated test benches for communication protocols running on Xilinx FPGA to catch edge cases and failure modes
• Participated in requirement analysis and verification
• Conducted rigorous code and requirement reviews

• Design Assurance Level A (DO-254) compliance, zero tolerance for undefined behavior
• Strict documentation and review processes
• Complex interaction between hardware and software components
• Early focus on simulation and verification to reduce late-stage risk
• Prioritized traceability between requirements, implementation and tests

Contributed to achieving quality certification for safety-critical aviation systems. Improved confidence in protocol correctness and system safety while reducing risk during integration phases. Demonstrated ability to work in regulated, safety-critical environments with strong discipline around quality, verification and documentation.

Independent researcher responsible for the entire research project, from problem definition to implementation and evaluation.

• Built complete ML pipeline starting with MNIST dataset for validation
• Worked with proprietary mmWave hand gesture datasets from Imec
• Implemented Multi-Class Gaussian Processes with scalable libraries
• Evaluated different kernel choices and approximation strategies
• Used CNNs as baseline for performance comparison
• Optimized for training time, inference speed and parameter count

• Working with high-dimensional, noisy radar data
• Balancing model accuracy with computational efficiency for edge deployment
• Limited compute and power budgets for target IoT devices
• Designing custom features from raw radar signals
• Working in Linux-based remote compute environments

Demonstrated that Gaussian Process models could achieve competitive performance with significantly fewer parameters than traditional approaches. Identified critical trade-offs between model expressiveness and computational efficiency. Showed the viability of non-deep-learning approaches for specific embedded ML use cases, providing insights relevant to energy-efficient edge AI systems. Demonstrated ability to work independently on open-ended technical problems and systems-level thinking.