Red Hat AI Model Deployment Workflow from Core to Edge
Project Overview
Using state-of-the-art computer vision and AI technology, we’ve developed a system that accurately detects defects in nuts as an example us case.
Archetecture Overview
The goal of this architecture is to streamline the process from model development to deployment, particularly in edge computing scenarios where models need to be run closer to data sources for faster processing. It emphasizes continuous integration and deployment (CI/CD) practices, automation, and the use of containerization for easy scalability and management. The model performance monitoring and training data collection at the bottom suggests that the system also includes feedback loops for continuous improvement of the AI models.
Configure Red Hat OpenShift Data Science
Explore how we configure Red Hat OpenShift Data Science for our AI model training. Read More
Jupyter Lab for Model Training
Download our Jupyter Lab notebook and explore how we train our AI models. Run notebook and deploy Pipeline
Deployment and Orchestration
Discover how we deploy our AI models using Microshift and manage workloads efficiently. Read More
Documentation for OpenShift Deployments
Check our detailed documentation for deploying on OpenShift.
Tekton Pipeline for CI/CD
We use Tekton Pipelines for continuous integration and continuous deployment. Run Pipeline
Microshift Deployment Instructions
- Configure DevSpaces for Ansible Automation Platform Configuration. Read More
- Configure Ansible Automation Platform. Read More
- Deploy our application on Microshift. Read More
Stay Connected
Follow us on GitHub and stay updated with our latest developments.
Posts
subscribe via RSS