lesterjessh.space

Lester Heyrana

Computer Engineer · Software Engineer

I’m a computer engineer working across software development, embedded systems, and machine learning. I’m currently the lead developer for GAKIT, a geohazard research project by MSU-IIT.

Explore my work ↓GitHubLinkedIn
01 / SELECTED WORK

Built with purpose.

From circuits to software, a selection of projects exploring how technology can make a difference.

GEOHAZARDS / DECISION SUPPORT01 / 08

GAKIT

Live

GAKIT supports geohazard assessment, flood simulation, weather, climate, and typhoon monitoring. It also collects community hazard reports and validates them automatically.

View details
TypeScriptPythonNext.jsFastAPIPostGISSupabaseONNX RuntimeDocker
WEATHER / GEOSPATIAL02 / 08

BAKAS

Live

Typhoon tracking with PAGASA advisories, Himawari imagery, and historical playback. Its tracking tools are integrated in GAKIT.

TypeScriptPythonNext.jsMapLibreSharp
GEOSPATIAL / ENVIRONMENT03 / 08

GeoSafe

Live

Landslide susceptibility mapping, satellite change analysis, and field observations. Its assessment tools are integrated in GAKIT.

TypeScriptPythonReactViteFastAPIMapLibreRasterioDocker
HARDWARE / MACHINE LEARNING04 / 08

Portable Spectrophotometer

A low-cost, portable spectrophotometer and machine learning platform designed for estimating heavy metal concentrations in water. Awarded Best Thesis and Academe Expo champion.

View design
PythonPyQt6scikit-learnPyTorchXGBoost
HARDWARE / EMBEDDED SYSTEMS05 / 08

3D Scanner

A 3D scanner with bare-metal AVR firmware, infrared sensing, stepper control, and SD card storage for captured point clouds.

View design
CAVRATmega328PFatFs
MACHINE LEARNING / POINT CLOUDS06 / 08

3D Shape Classifier

A PointNet model that classifies scanned point clouds into 3D shapes, with tools for training, evaluation, and interactive inference.

PythonPyTorchPointNet
DIGITAL DESIGN / FPGA07 / 08

32-bit RISC Microprocessor

A processor built in Verilog on a DE1-SoC board, including registers, ALU, memory, and control unit.

VerilogDE1-SoC
EDGE AI / EMBEDDED08 / 08

Drowsiness Detection

A lightweight CNN and embedded detection algorithm designed to run on the low-power MAX78000 microcontroller.

CPythonMAX78000
More on GitHub ↗
02 / ABOUT

A bit about me.

I’m based in the Philippines and studied Computer Engineering at MSU-IIT.

My work includes scientific instruments, geospatial applications, and on-device AI. I also take on freelance software development projects.

LANGUAGES

Python · TypeScript · C / C++ · PHP · Verilog

FRAMEWORKS & TOOLS

React · Next.js · FastAPI · PyQt6 · Docker

ML & GEOSPATIAL

PyTorch · scikit-learn · ONNX Runtime · Rasterio · PostGIS

EMBEDDED & HARDWARE

Arduino · MAX78000 · FPGA · PCB Design

03 / SAY HELLO

Have an idea?
Let’s build it.

Email me ↗