Devops for machine learning models
Modops supports the deployment and monitoring of machine learning models in a simple and self-service manner and exposes a REST API to your models that’s instantly consumable by your development team.
Compatible with third party models
ModOps works with models built on Scikit-Learn, R, Apache Spark and, of course, Learner. You can also import models encoded in PMML files.
REST API access to your models
Your development team just needs to know the business data to pass to the API to make predictions.
Operational and performance insights
We believe the same philosophy and rigor that goes into releasing software should be applied to model deployment.
Model Lifecycle Management
Get timely performance alerts and know when to retrain your model so it doesn’t get stale.
Detailed reporting on predictions
Delve deeper into your predictions by blending in external data that goes beyond the features of your model.
Works with Teraport and Learner
Opt in to model retraining with high quality production data used in making predictions.