Skip to content
View dmazilkin's full-sized avatar

Block or report dmazilkin

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
dmazilkin/README.md

Education:

  • Bachelor's degree: Mechatronics, Brno University of Technology, 2021-2024.
  • Master's degree: Applied Informatics, Brno University of Technology, 2024-2026.

Specializations:

  • Embedded systems,
  • Machine Learning, Deep Learning, Data Science.

Tech stack

Embedded systems:

C C++

Machine Learning, Deep Learning, Data Science:

Python

Work Experience

Embedded Software Engineer
Bender Robotics. May 2023 — December 2024.

  • Firmware development for STM32 Cortex M in compliance with MISRA C.
  • Development under FreeRTOS, HAL and Bare Metal.
  • Build system setup using CMake, Makefile.
  • Peripheral integration: ADC, DAC, UART, USB, I2C and SPI.
  • Communication systems using HDLC and Protocol Buffers.
  • Unit testing using Unity Test Framework.
  • Code versioning and collaboration via GitLab.
  • Maintained a CI/CD pipeline with linters and static analysis tools.

Key Achievement:

  • Fully designed and implemented from scratch a modular firmware and Python PC API for the company's Touchscreen Tester device.
  • Automated touchscreen testing on the production line, reducing manual effort and improving test consistency.

Variant Management Tool Developer
NXP Semiconductors. February 2025 — Present.

  • R&D in Product Line Engineering following the V-model development process.
  • Development of Feature Models and Variants using pure::variants and Eclipse.
  • Integration of IBM DOORS Next and IBM ELM RM to automate the flow of requirements into variant models.
  • Automation of configurations via pure::variants JavaScript API, Eclipse console, CMake, Makefile and Python.
  • Task and code management using Jira and Bitbucket.

Personal Projects

Algotihms and Data Structures Projects

Maze Pathfinding with Search Algorithms – Graph Search and Route Planning - Implemented classic pathfinding algorithms (BFS, DFS, A* and Jump Point Search) to solve mazes represented as grids. Implemented and Used efficient data structures: Double-Linked List for stacks/queues and Binary Heap for priority queues. Visualized results and stored paths and performance metrics using Pandas and image exports using Pillow.

File Compression with Canonical Huffman Code – Data Compression Algorithm - Implemented Canonical Huffman coding in C for compressing and decompressing ASCII files. Built frequency table and Huffman tree using a min-heap implemented from scratch, generated Canonical Huffman codes, and saved metadata for decompression. Developed a command-line argument parser to handle compression and decompression options with input/output file paths. Compiled the project using Makefile and processed files in chunks to handle large data efficiently.

TSP Solver – Metaheuristic Algorithms for the Traveling Salesman Problem - Implemented Genetic Algorithm and Simulated Annealing in C++ for solving the TSP. Applied Object-Oriented Programming and the Strategy Design Pattern to enable modular algorithm selection via a common interface. In Simulated Annealing, used swap for neighbor generation at high temperatures and 2-opt at lower temperatures for better refinement. Designed a flexible architecture with a TSPContext class managing solver strategies and an abstract TSPSolver base class. Supported specific configurable runs, file-based input/output, and built the project with a Makefile.

Machine Learning Projects

Machine Learning algorithms implemented from scratch: Linear Regression, Logistic Regression.

Function Approximation and Deployment with Neural Network – Regression - Trained a neural network to approximate a mathematical function using synthetic data. Saved the model and preprocessing parameters in JSON for loading model. Deployed the model with FastAPI and Uvicorn, exposing a REST API for predictions. Evaluated the model's performance using MSE, MAE and metrics.

Data Classification with Neural Network - Classification - Built and trained a binary classifier using Keras on a medical dataset. Saved the model and preprocessing parameters to JSON for loading and using pretrained model for inference. Evaluated the model's performance using Accuracy, Precision, Recall, F2Score and Precision-Recall Curve metrics.


📫 Contacts

Pinned Loading

  1. Compression-Huffman-code Compression-Huffman-code Public

    Canonical Huffman compression algorithm from scratch.

    C

  2. NN-Regression-WebApp NN-Regression-WebApp Public

    Neural network for Regression problem.

    Jupyter Notebook

  3. TSP-Solver TSP-Solver Public

    Travelling Salesman Problem Solver implemented in C++.

    C++

  4. ML-LinearRegression ML-LinearRegression Public

    Linear Regression from scratch.

    Jupyter Notebook

  5. ML-MazeSolver-RoutePlanning ML-MazeSolver-RoutePlanning Public

    This repo contains maze solver, which is implemented with DFS, BFS, A* and JPS algorithms.

    Python

  6. ML-LogisticRegression ML-LogisticRegression Public

    Logistic Regression from scratch.

    Python