C.A.R.A (i)
An early AI assistant experiment started around 2013, combining speech recognition with a dynamically learning language library. The goal was to interact in a more open, conversational way instead of using fixed commands.
About me
Short background and where C.A.R.A started.
I’m a developer who enjoys experimenting with unusual combinations of tools and technologies. Over the years that has led me from small utilities and experiments to full engines and AI prototypes.
Around 2013 I started working on C.A.R.A (i), a relatively simple AI at first. I began with speech recognition, then added a dynamically learning language library, and over time it became possible to have fairly natural conversations with it. C.A.R.A no longer exists in its original form, but the project strongly influenced how I think about language, interaction and tools today.
Selected Projects
A few things I’ve built or experimented with.
NeuralNetwork
A lightweight neural network implementation written in C#. It focuses on being straightforward to integrate into .NET projects, making it useful for small machine learning experiments without relying on large external frameworks.
View on GitHub →Plain-GDI-3D
A prototype 3D engine built on top of classic GDI+ drawing. It renders basic 3D scenes using only 2D drawing primitives, mainly as a playground for engine concepts and low-level rendering without a dedicated 3D framework.
View on GitHub →