A Machine Learning solution with a small footprint and vast possibilities.

Seamless Integration. Local Execution Without GPU Requirements. Respects privacy of your users.

Next generation machine learning solution to make your software smarter without bloat.

Easy to Integrate

Build model as a native or WebAssembly library, and integrate with your existing systems.

Wide Range of Platforms

Native libraries with the support for most modern platforms. WebAssembly support for running in a web browser.

Ease of Deployment

No additional dependencies. No need to change your existing deployment process.

Fast and Efficient

Small download size. Low memory usage.

For Software Developers

Leverages best software development practices. No need to add new tools or languages.


Native code. Uses your CPUs to its full potential. No garbage collection pauses.


    Uchen.ML Runtime

    Experience high-performance and efficient model inference with Uchen.ML Runtime, a C++20 runtime that integrates seamlessly with modern platforms with compatible compilers.

    Uchen.ML Training Core

    A C++20 library and a collection of command-line tools for training machine learning models. Supports common workflows with a particular focus on reinforcement learning algorithms.

    Uchen.ML Console

    A hosted tool for quick onboarding and managing Uchen models. Intuitive interface for managing model versions, datasets and deployment targets.

    Uchen.ML Brain Surgeon

    A hosted tool for debugging and refactoring Uchen models. It offers deep insights into how the model's "thinks" and encourages teams to explore new ideas.