- Deployment pipelines are automated
- Data pipelines are neatly integrated for filtering, masking and cleansing
- Data is prepared periodically for a new training round
- Manage configurations, resources, and provisioning for training and production deployment
- Setting up tracking and versioning for experiments and model training runs
- Setting up the deployment and monitoring pipelines for the models that do get to production
- Incident management and response
- Assisting end-users in using the application and answering their queries
- Monitoring and alerting for problems and performance issues
- Deploying centralized logging
- Troubleshooting for deployment, capacity, connectivity, resources, and function
- Management of applications & platform configurations
- Coordination of changes and service requests
- Feedback to development teams for application bugs and glitches
- Application documentation