C • Low-level systems • Graphics

Building closer
to the metal.

I'm Nick Bedner, a C programmer focused on systems, graphics, and performance-critical software. I like understanding how things work all the way down.

Featured project / 01

Driving-ML

A custom driving simulation and reinforcement learning experiment, built with a low-level engine in C and assembly.

Engineered from the ground up.

I designed and developed a custom engine in C and assembly without external libraries, implementing rendering across Vulkan, DirectX 12, and Metal. The project combines manual memory management, graphics programming, and reinforcement learning in a driving environment.

CAssemblyVulkanDirectX 12MetalReinforcement Learning
Technical toolkit

The tools I build with.

My work is rooted in C, assembly, graphics APIs, and understanding software at a low level. I also bring experience in networked applications, web development, and performance-focused engineering.

C
Assembly
Vulkan
DirectX 12
Metal
OpenGL
Java
JavaScript
Multithreading
Memory Management
Networking
HTML / CSS
DirectX 11
Graphics Programming
Reinforcement Learning
RabbitMQ
Database Synchronization
Low-Level Systems
CNC Programming

Languages, graphics APIs, and systems technologies I have worked with.

The background

Experience & education

2019 — Present

CNC Machinist / Precision Grinder

Self-employed. CNC programming, precision manufacturing, technical schematics, and hands-on troubleshooting.

Feb — Oct 2019

Junior Software Engineer

Perspectium. Java and JavaScript development, network programming, database synchronization, RabbitMQ, and web technologies.

2025 — Present

California State University San Marcos

Computer Science degree in progress.

2025 — Present

Palomar College

Guitar Making Technology and Cabinetry certificates in progress. An interest in making things extends beyond software.

About me

Software meets craftsmanship.

My work spans software engineering and precision machining. Whether I'm writing graphics code or programming CNC equipment, I'm drawn to the details: performance, accuracy, and the satisfaction of building something that works.

Enlarged Driving-ML screenshot