Waveshaper AI Inference Engine

February 13, 2024 ยท One minute read

In this blog post we talk about Waveshaper AI’s Inference Engine and how it can provide codeless deployment.

Waveshaper’s Inference Engine is a key component of our comprehensive end-to-end framework designed to address a wide array of audio problems. This versatility enables us to tackle challenges such as spectral recovery, various audio enhancements, modelling analog devices, denoising, with remarkable efficiency and effectiveness.

The core of our inference engine is written in highly optimized C++ code, ensuring compatibility and consistent output with leading AI training frameworks like PyTorch and TensorFlow. This optimization allows our engine to operate in real-time across a broad spectrum of hardware platforms. It’s not just a tool; it’s the backbone of our Software-as-a-Service (SaaS) offerings and plugins, including VST and AU formats.

Effortless Functionality Updates

One of the standout features of our inference engine is its codeless deployment capability. Imagine it as a versatile “game console” for audio processing. Once the inference engine is set up on a hardware platform, updating or changing its functionality is as simple as swapping game cartridges - no recompiling necessary. This approach significantly reduces development time and resource allocation, making it an extremely agile solution in a fast-paced industry.