In their re:Invent 2019, Las Vegas conference, Amazon made a series of an announcement on its new products related to AI and IoT. The tech giant introduced their AI-enabled Piano keyboard called Deepcomposer, which is based on Amazon’s Transcribe Medical, a new edition of the transcribe speech recognition service. This enables the developers to add speech-related data, which develops the speech to text capabilities in the apps. The DeepComposer allows the users of AWS (Amazon Web Service) to create music using AI and a MIDI controller.
The transcribe medical of Amazon is covered under HIPAA eligibility and Business Associate Addendum (BAA). Other than Amazon, tech giants like Microsoft and Philips are also working on speech recognition AI software.
The keyboard is a 32 key, 2-octave keyboard which is designed to give the developers the experience in Generative AI where complex coding and programming are not involved. The composers can record a musical tune or use a pre-recorded tune before selecting the model for their genre and other parameters. After this, they can choose the hyperparameters and a sample for validation. The DeepComposer then produces the composition which can be played in the AWS console. This data can also be exported to a sound cloud. The main aim of DeepComposer is to get the consumers to involve directly in Machine Learning.
The main aim of this medical transcription technology is to enable the doctors, nurses, and researchers to enter the clinical data to the system using AI, thus improving the speed and efficiency of data entry, saving up to 6 hours of labor time.
AWS has also launched Amazon SageMaker Operators for Kubernetes. With this, the data scientists using Kubernetes can train, tune, and deploy AI models in the Sagemaker ML platform of Amazon. The Sagemaker operators can be installed in Kubernetes clusters by customers to create Sagemaker jobs using Kubernetes API and command-line tools.
According to Aditya Bindal, AWS Deeplearning product manager, “Now with Amazon Sagemaker Operator for Kubernete, the customers can continue to enjoy the portability and standardization benefits of Kubernetes… along with integrating the many additional benefits that come out of the box with Amazon Sagemaker, no custom code is required”.
To support ML workloads, however, the customers have to write custom codes to optimize ML infrastructure, ensure high-reliability AD availability, and comply with appropriate regulatory and security requirements.