Cerence AI has expanded its collaboration with NVIDIA to enhance the capabilities of its CaLLM™ family of language models, which includes cloud-based and embedded forms for automotive applications. By utilizing the NVIDIA AI Enterprise software platform, specifically the NVIDIA TensorRT-LLM library and NeMo framework, Cerence aims to optimize its language models for faster in-vehicle assistant performance and improved user experience. This collaboration allows Cerence to leverage NVIDIA’s technology to accelerate the deployment of AI solutions, ensuring better interaction and privacy in vehicles. Cerence has implemented specialized features like NVIDIA NeMo Guardrails and agentic architecture on its CaLLM Edge models to advance in-vehicle user experiences. Nils Schanz of Cerence highlights the benefits of improved cost savings and performance for automakers, enhancing driver safety and productivity.

Rishi Dhall of NVIDIA emphasizes the challenges of deploying large language models due to their complexity, which this partnership helps address. Cerence Inc. is a leader in AI-powered automotive experiences, working with major automakers globally to enhance user experience and vehicle interaction. The company is based in Burlington, Massachusetts, and continues to push boundaries in AI innovation. For more information about Cerence AI and its advancements, visit their website or follow them on LinkedIn.