Computer Engineering graduate (University of Victoria) and Engineer-in-Training at Venturi Engineering Solutions, where I work on the simulation and data systems behind pipeline leak detection and operator training. I'm most interested in work where data, modelling, and software meet: machine learning, estimation, and the tooling that keeps real-world data trustworthy.
🌐 Portfolio: benphilipenko.ca · 💼 LinkedIn: benjamin-philipenko
| Project | What it is | Stack |
|---|---|---|
| Adaptive Machine Learning for Radiolocation | Parzen-window position estimation from WiFi signal strength, fused with an adaptive Kalman filter to track a moving target (cuts tracking error by about two-thirds and beats a neural-network baseline) | MATLAB |
| Re-Bike | Senior capstone (team of 5): regenerative-braking e-bike conversion. I led the user interface subsystem: a Teensy controller that handles throttle and brake input, sends motor commands over CAN bus, and drives an OLED dashboard | C++, CAN bus |
| Antarctic Vision | YOLOv5 object detection and segmentation, with a custom 1,600-image dataset to detect penguins | Python, PyTorch, Colab |
| NHL Fantasy Score Predictor | Weighted-average and linear-regression ensemble predicting fantasy scores from 15 seasons of NHL data | Python |
| Leviathan | Low-cost ocean-floor mapping prototype (Raspberry Pi, sonar, GPS) with interactive 3D terrain maps, backed by an Ocean Wise Ocean Action Grant | Python, Node.js, Plotly |
| Embedded PWM Control | Bare-metal STM32 system: interrupt-driven frequency measurement, ADC/DAC closed-loop control of a 555 timer, custom SPI OLED driver | C, ARM Cortex-M0 |
| Personal Portfolio Website | This portfolio: a responsive single-page app | Vue.js, Vuetify, Firebase |
Python · MATLAB · C/C++ · TypeScript/JavaScript · Vue.js · SQL · PowerShell · Git · SCADA & leak detection systems · hydraulic simulation
