About Me
Embedded Systems Engineer with 5+ years of commercial experience in low-level C/C++ firmware development, secure hardware-software integration, and real-time system diagnostics for production-grade embedded devices.
Currently pursuing Master's in Robotics and Control Systems at Gdańsk University of Technology (GPA 4.46/5.0), I combine deep embedded Linux expertise with advanced research in machine learning, signal processing, and computer vision for autonomous systems.
My professional background spans bootloader development, device driver programming, JTAG debugging, and security implementation for high-volume consumer devices based on Broadcom SoCs. I excel at bridging hardware constraints with software innovation to deliver robust, scalable embedded solutions.
Firmware Development
Bootloaders, device drivers, interrupt handling, real-time constraints
Security & Encryption
AES encryption, secure boot, OPTEE, RPMB, cryptographic key management
Hardware Integration
UART, I2C, SPI, GPIO, JTAG debugging, schematic analysis
Autonomous Systems
Robotics, sensor fusion, path planning, machine learning integration
Featured Projects
A selection of academic and personal projects demonstrating embedded systems expertise, real-time programming, and autonomous system development.
Autonomous Mobile Robot Co-Simulation (CarryBot)
Completed • Feb 2026Airport logistics robot with "split-brain" architecture — decoupling high-level Python AI (A* pathfinding, YOLOv8 computer vision) from low-level Arduino embedded safety protocols communicating via UART.
Arduino Mega 2560 firmware implementing state machine logic (LOCKED/UNLOCKED), 15cm ultrasonic safety envelope, IMU-based heading control (MPU6050), and emergency stop protocols.
Integrated Webots R2025a physics simulation with Proteus 8 circuit verification, validating embedded firmware before hardware deployment. Achieved 48ms control latency for real-time telemetry.
Built operator interface using JavaScript (Flask-SocketIO, Chart.js) with live camera feeds, battery voltage monitoring, GPS position tracking, and bi-directional command/control interface.
Implemented A* pathfinding on 0.2m grid, YOLOv8-based pedestrian detection, and lidar-based obstacle avoidance with real-time sensor fusion.
Key Achievements
- Architecture: Successfully decoupled mission-critical safety logic from high-level AI, applicable to automotive, robotics, and aerospace domains
- Testing: Hardware-in-the-loop methodology reduced hardware iteration costs and validated firmware reliability pre-deployment
- Integration: Seamless UART communication protocol between Python supervisor and Arduino safety controller
Cognitive Architecture for Autonomous Agents
Ongoing • Master's ThesisResearch Project: Validating that autonomous agents with episodic memory and subjective qualia layers exhibit superior adaptive behavior in novel hazardous environments compared to standard reactive agents — implemented in Webots R2025a factory rescue simulation.
Developed CUDA-accelerated YOLOv8 object detection, K-means color extraction, and HSV hazard detection modules with BLAS-optimized cosine distance metrics for real-time feature vector comparison.
PyTorch CNN/LSTM episodic memory model generating κ (kappa) traces for fuzzy-logic decision-making, targeting sub-100ms decision latency.
FastAPI + WebSocket dashboard with Chart.js visualization showing live agent states, annotated camera feeds, and performance overlays (FPS, GPU utilization, memory usage).
Implemented Action Potential formula combining episodic memory confidence with Universal Fear baseline for context-aware hazard response.
Applications
- Autonomous Vehicles: Decision-making in GPS-denied or unmapped environments
- Robotics: Adaptive behavior learning for search-and-rescue operations
- Industrial Automation: Fault-tolerant navigation in dynamic factory floors
Environmental Sound Classification System
Ongoing • Updated March 2026ML-powered acoustic warning system for autonomous agents to classify 5 safety-critical environmental hazards in real-time: emergency sirens, vehicle collisions, train horns, human screams, jackhammer sounds.
Processed 1,000+ raw 44.1kHz mono audio samples into Log-Mel Spectrograms using Short-Time Fourier Transform (STFT). Implemented audio augmentation (pitch shift, time stretch, noise injection) for robust generalization.
Achieved 92% classification accuracy with sub-100ms inference latency on dual-source dataset (Freesound + UrbanSound8K).
Built FastAPI REST backend serving model predictions alongside Streamlit web UI for audio upload, real-time classification, and confidence visualization.
Merged and preprocessed data from two independent sources, handled class imbalance, created 3 stratified train/validation splits for cross-validation.
Industrial Applications
- Automotive: Acoustic hazard detection for ADAS and autonomous vehicles
- Robotics: Environmental awareness for mobile robots in industrial settings
- IoT/Edge Devices: Real-time audio classification on resource-constrained hardware
- Safety Systems: Workplace hazard monitoring and emergency response triggering
AWS Weather Trends Pipeline
Completed • June 2025End-to-end cloud data engineering pipeline that processes historical weather data using Apache Spark (PySpark) on an AWS EMR cluster, stores results in Amazon S3, generates trend visualizations, and forecasts future weather conditions using a custom-built linear regression model.
Configured and deployed a multi-node EMR cluster (emr-7.9.0) with Spark 3.5.5, Hadoop 3.4.1, and Hive on m4.large EC2 instances. Processed raw weather CSV data from S3 using PySpark with S3A connector for scalable data ingestion.
Computed weekly and monthly aggregates (min/max/avg temperature, wind speed, total precipitation) using PySpark groupBy operations with proper data type casting and null handling for data quality.
Built a least-squares linear regression model from scratch without external ML libraries. Generates 4-week ahead forecasts for temperature, wind speed, and precipitation with a max(0, prediction) constraint to prevent physically impossible negative values.
Designed S3 bucket architecture for data lifecycle management. Configured IAM roles (EMR_DefaultRole, EMR_EC2_DefaultRole) for secure access. Automated plot generation and S3 upload via Boto3. Integrated GitHub via SSH from EMR master node using PuTTY.
Key Achievements
- Architecture: End-to-end pipeline: S3 (data) → EMR (processing) → S3 (output) with automated visualization and forecasting
- AWS Skills: Hands-on experience with EMR cluster provisioning, S3 storage management, EC2 instances, and IAM role configuration
- DevOps: SSH key-based GitHub integration from EMR master node, spark-submit job execution, and reproducible cloud workflows
Additional Projects
VitalNexus - Patient Telemetry System
Real-time medical telemetry dashboard with ECG waveform visualization, alarm intelligence, and multi-patient monitoring for 4-8 concurrent patients.
Impulsive Noise Removal
Adaptive AR model with exponentially weighted least squares (EW-LS) and linear interpolation for removing impulsive noise from music recordings.
Emotion-Aware Perception Module
Real-time computer vision module analyzing crowd flow to provide autonomous agents with high-level perceptual intelligence. Collaborative research project.
Access to Private Repositories
Several projects are hosted in private repositories due to ongoing research or collaborative work under institutional guidelines. I'm happy to provide:
- Code Access: GitHub collaborator invitations for technical review
- Documentation: Detailed technical reports, architecture diagrams, and API specifications
- Demonstrations: Recorded demos, live walkthroughs, or deployment access
Professional Experience
Software Engineer
Key developer for AT&T Residential Gateway project (BGW210, BGW320, CGW450) based on Broadcom chipsets, focusing on SDK migration, bootloader development, and security implementation for 5M+ deployed devices.
Core Responsibilities & Achievements:
-
Embedded Linux Development:
Authored and optimized production-grade low-level C code for aarch64 systems. Developed bootloaders (U-Boot/CFE), device tree configurations, and memory management modules for high-volume consumer devices.
-
Security Architecture Implementation:
Implemented AES-256 encryption during compilation, integrated SecureOS for OPTEE (Trusted Execution Environment), RPMB (Replay Protected Memory Block), and ROOTFS encryption, ensuring strict data execution and tamper-resistant storage.
-
Device Driver & Peripheral Integration:
Developed and verified low-level controller functionality for 5+ core hardware peripherals: GPIO, SFP modules, LEDs, Watchdog timers, EMMC storage. Proficient in interrupt handling and character device driver development.
-
Platform Validation & Debugging:
Utilized JTAG debuggers (Lauterbach, OpenOCD) to diagnose and resolve critical platform bugs. Achieved 98% first-pass hardware validation success rate. Confirmed BIST (Built-In Self-Test) functionality across 3 hardware generations.
-
Hardware-Software Co-Development:
Interpreted complex hardware schematics and datasheets for Broadcom SoC platforms (BCM6858, BCM6813, BCM63148). Ensured reliable hardware-software integration during SDK migrations and DDR interface validation.
-
DevOps & Automation:
Established Docker-based CI/CD pipelines for reproducible cross-compilation. Reduced deployment errors by 35%. Implemented unit testing frameworks, reducing field defects by 40%.
Software Engineer Intern
Gained foundational experience in embedded C development lifecycle, code review processes, and debugging for bootloader components.
- Assisted senior developers with entry-level C programming tasks and low-level system initialization routines
- Developed Python automation scripts for hardware regression testing workflows
- Participated in Agile sprint planning, code reviews, and technical documentation
Recognition & Awards
Individual Excellence Award
Recognized for demonstrating accountability, technical leadership, and exceptional performance during annual reviews at Capgemini.
Analytical Skill Commendation
Commended for rapid assimilation of new modules and strong analytical capabilities in addressing field-related critical issues.
Technical Skills
Programming Languages
Embedded Systems & RTOS
Hardware Integration & Debugging
Signal Processing & Sensor Fusion
Security & Encryption
DevOps & Development Tools
Autonomous Systems & Machine Learning
Education
Master of Science in Automatic Control, Cybernetics, and Robotics
Specialization: Intelligent Decision Systems and Robotics
Bachelor of Engineering in Electronics and Communication
Foundation in: Microprocessors & Microcontrollers, Digital Signal Processing, Communication Systems, VLSI Design, Control Systems Engineering
Let's Connect
I'm actively seeking opportunities in embedded systems development, firmware engineering, robotics, IoT, and real-time systems. Available for remote work or relocation within EU.
Interested in discussing opportunities or collaboration?
Feel free to reach out — I'm always open to interesting projects in embedded systems,
robotics, and autonomous technologies.