Skip to content
View JJKSweaty's full-sized avatar

Highlights

  • Pro

Block or report JJKSweaty

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.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

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

Report abuse
JJKSweaty/README.md

Hi, I'm Jonathan Koshy

Electrical Engineering @ University of Waterloo | Firmware, Embedded Systems & Edge AI

PortfolioEmailLinkedInGitHub


About Me

I design, build, and validate embedded systems that connect firmware, sensors, real-time control, and hardware. My work focuses on low-level drivers, RTOS task pipelines, communication protocols, PCB bring-up, and hardware/software integration under tight timing and memory constraints.


Technical Focus

Image Image Image Image Image Image Image Image Image


What I Work On

  • Embedded firmware in C and C++ for microcontrollers
  • Bare-metal firmware, memory optimization, and peripheral bring-up
  • FreeRTOS task design, scheduling, and real-time data pipelines
  • Sensor drivers and biosensor data acquisition systems
  • Communication over BLE, Wi-Fi, UART, I2C, SPI, CAN, and TCP/IP
  • Embedded UI development using LVGL
  • Hardware bring-up, oscilloscope debugging, logic analyzer validation, and PCB testing
  • RTL design in Verilog for digital systems and packet parsing

Edge Computing and On-Device Perception

I’m actively exploring edge computing for real-time perception and control, with a focus on running vision workloads close to the hardware to reduce latency and avoid cloud dependencies.

My recent work includes building an edge AI assistant on a Raspberry Pi 5 with a Hailo-8L accelerator, using YOLOv8 and OpenCV for real-time object detection, medication verification, and persistent object memory. I’m interested in how embedded Linux systems, AI accelerators, and microcontroller firmware can work together in systems that sense, decide, and actuate in real time.

This connects directly with the kind of embedded systems I like building: real hardware, real constraints, and real testing.

Projects

Detailed project write-ups, system diagrams, images, and demo videos are available on my portfolio site: https://jonathankoshy.netlify.app

Featured work includes:

  • AI Dementia Companion — Raspberry Pi 5, Hailo-8L, YOLOv8, OpenCV, TypeScript
  • High Speed Autonomous Disk Launcher — Raspberry Pi 5, ESP32-S3, OpenCV, PID control, LiDAR
  • ESP32 Wi-Fi Media Controller — ESP32, LVGL, Wi-Fi, Spotify API, embedded touchscreen UI
  • RISC-V Interrupt Driven Embedded Controller — SiFive FE310, RISC-V assembly, GPIO interrupts
  • ASIC Ethernet Parser — Verilog, Ethernet frame parsing, ARP validation, hardware packet processing

Collaboration

I’m open to collaborating on embedded systems, robotics, firmware, edge AI, and hardware/software projects involving real devices, real constraints, and real testing.


📫 Contact

Email: jjacobko@uwaterloo.ca Portfolio: https://jonathankoshy.com Location: Canada

Pinned Loading

  1. efs-zeropilot-4.0 efs-zeropilot-4.0 Public

    Forked from UWARG/efs-zeropilot-4.0

    C 1

  2. tcpip-stack tcpip-stack Public

    TCP/IP stack in C for Linux, built step-by-step from Ethernet and ARP to IPv4, ICMP, TCP handshakes, sockets, and retransmission.

    C 1

  3. ESP32Media ESP32Media Public

    ESP32-based dashboard designed for media playback monitoring, process management, and discord controls

    C++ 1

  4. MNIST MNIST Public

    2-layer MLPfor MNIST digit classification

    Cuda 1

  5. jjkVR jjkVR Public

    Custom 2K 120Hz VR headset with STM32F411, 9-DoF IMU tracking, custom PCB, and SteamVR support.

    C 2