Monitor deployed models for accuracy, drift, and performance and quickly retrain to reduce risk and ensure business value. Track models in a single pane of glass.
Deploy to scalable infrastructure across both on-premise, any cloud, or hybrid environments to ensure high-performance inference and responsive models and apps.
Domino supports your deployment strategy and infrastructure. Deploy in Domino, in-database, the edge, hosted services like Sagemaker or to existing CI/CD pipelines.
Model Registry
Track all your models regardless of where they were trained. With Domino Model Registry, you get complete lineage tracking for auditability using integrated model cards. The model registry offers a central repository of all models, streamlines iterative improvement, and facilitates stakeholder reviews and approvals for transitioning models from development to production.
Model Review and Validation
Validate and review models with custom approval workflows to ensure models and applications are robust and audit-ready with best-in-class reproducibility. Provide reviewers with detailed lineage and model metrics to help them evaluate AI trustworthiness.
One-click deployment
Deploy models to any endpoint for both batch and real-time predictions. Deploy in Domino natively, integrate with existing CI/CD pipelines, or export models to platforms like SageMaker, Snowflake, Databricks, or NVIDIA FleetCommand. Deploy in a hybrid world — on-prem, across multiple on-prem environments, or a combination of on-prem and cloud. Share analytical dashboards, AI models, and GenAI apps with any framework, including Dash, Flask, Streamlit, Shiny, and more.
Model Monitoring
Track data drift and model quality degradation automatically with integrated monitoring and alerts. Continuously monitor accuracy metrics and ground truth to improve performance. Incorporate LLM evaluation frameworks. Monitor endpoint activity and health with prebuilt or custom metrics. Quickly identify and remediate issues and retrain with ease.
