Automation technology to develop, operate, and improve AI
An AI technology operation system which can seamlessly handle development, operation, and improvement
Technology of Konica Minolta, which quickly deploys AI services
Konica Minolta can quickly deploy AI services and continuously improve its AI model.
We are developing efficient and reliable technologies to implement the entire process, such as strategy management of AI services, data collection, automated training and continuous improvement of AI models.
MLOps is a term that combines “Machine Learning” and “Operations”.
We are developing systems that can maintain high reliability even in the customer’s environment where the situation is always changing. This is made possible by providing a framework that automatically implements the cycle of AI model development, operation, and continuous improvement.
An automated continuous integration system for AI models
We use a Git service to automate the testing and analysis of developed AI models. This technology enables rapid AI model updates and quality assurance in various development and operating environments (e.g. Windows/Linux, cloud/on-site).
Centralized management system for evaluating AI models
To maintain the quality of AI models, it is important to assure the quality of datasets and AI algorithms. However, these are usually incomplete in the development phase, so data additions and model revisions are frequently repeated for actual environments. For efficient quality control, Konica Minolta can provide a system that centrally manages not only data and AI model versions, but also datasets, parameters and evaluation results used in experiments. With this system, we ensure the reproducibility of AI models and enable customers to securely use our AI services.