Groq is a provider of low-latency AI inference solutions. It offers an LPU Inference Engine, an end-to-end processing unit system that provides inference for computationally intensive applications with a sequential component, such as AI language applications. The solution can be used in audio, speech, image manipulation, video, scientific research, coding, etc.