We implemented a kubeflow pipeline specifically for a kserve application tailered for IOT Rasberry PI chipset. The pipeline takes in a regular input data, be a text or image, and feed into a trained model from other AI training processes. A node-red frontend is constructed with a custom node which can trigger the execution of the kubeflow kserve pipeline. The pipeline will takes an input from Rasberry PI sensor, and deliver an output back to the PI chipset for proper response such as playing a audio or beeping a buzzer.
Please let us know if this feature can be of interest to your group and where we should PR the code.
We implemented a kubeflow pipeline specifically for a kserve application tailered for IOT Rasberry PI chipset. The pipeline takes in a regular input data, be a text or image, and feed into a trained model from other AI training processes. A node-red frontend is constructed with a custom node which can trigger the execution of the kubeflow kserve pipeline. The pipeline will takes an input from Rasberry PI sensor, and deliver an output back to the PI chipset for proper response such as playing a audio or beeping a buzzer. Please let us know if this feature can be of interest to your group and where we should PR the code.