Automated authoring enables simplified deployment of applications and services for complex use cases, especially in the field of machine learning. This paper presents the development and implementation of a specialized authoring tool that can be used for computer vision applications, enabling automated creation of machine learning services. The proposed authoring tool realizes a microservices architecture to facilitate the conversion and deployment of machine learning inference services, especially in image classification and object detection use cases. The authoring process addresses the interoperability issues commonly faced in machine learning frameworks, leveraging the Open Neural Network Exchange (ONNX) for model conversion into a standardized format. By encapsulating machine learning tools in containerized applications, this authoring tool offers a modular solution that can be easily adapted to various industrial applications. The developed authoring tool integrates the common machine learning frameworks PyTorch and TensorFlow, coupling DevOps methodologies such as CI/CD, ensuring a robust, maintainable, and user-friendly system that meets the growing needs of machine learning use cases in manufacturing.