Model deployment & management
SAS Model Manager
Build a streamlined, secure ModelOps process.
Connect data scientists, MLOPs engineers and business analysts. Deploy models quickly. And integrate with open source.
Ensure model governance and transparency.
A centralized, searchable repository for all types of models and analytical assets gives you complete visibility into your analytical processes, ensuring traceability and governance. The solution simplifies model management with version control. You can track project history through each step of the model management process and get a unified view of each model’s currency, definition and value. Using open REST APIs to access models and model-score artifacts streamlines IT work.
Easily validate models to ensure high-quality predictions.
Use our open source package, sasctl, to automatically generate executable scoring code for Python-based models. You can easily test models, validating model scoring logic before models are pushed into production, from an easy-to-use no-code interface.
Build once, deploy everywhere – no recoding required.
Efficiently move your analytical models from the innovation lab into your chosen production environment. SAS Model Manager has you covered, whether your need to: deploy models into databases; score data in batch; host a real-time REST API scoring endpoint; push models into a container hosted in registries on Docker, Azure, GCP or AWS; or deploy directly into Azure Machine Learning.
Automatically monitor model performance to keep them performing as expected.
SAS Model Manager automatically monitors the performance of models – from inception, to usage, to retirement – regardless of the language used to create them. Performance benchmarking reports display models’ scoring performance and document conformity to required standards. Alerts are generated to indicate model decay. As models are used across different departments, the solution produces extensive tracking, validation and auditing reports, and marks champion models for use in other applications. Ongoing monitoring lets you know when it’s time to refine or retire a model.
Increase efficiency by adapting models to reflect internal or external changes.
Continuously update models to keep pace with changing market and business conditions. You can retrain the existing model on new data, or revise the model using feature engineering or new data elements. Model retraining integrates with the model pipeline processing environment for greater efficiency.
Save time and resources by automating the model life cycle using a CI/CD approach.
SAS Model Manager enables you to integrate multiple environments, tools and applications using open REST APIs. You can automate the analytic life cycle by creating custom workflows that match your business requirements and processes.
Free Trial
Try the latest SAS Viya capabilities.
Get a free, 14-day trial of SAS Viya, which includes SAS Model Manager, as well as capabilities for the entire analytics life cycle.
Key Features
Simplify model collection creation and management with a web-based interface that easily automates the model management process.
Data access, preparation & quality
Access, profile, cleanse and transform data with an intuitive interface that provides self-service data preparation capabilities with embedded AI.
Custom chatbot creation
Create and deploy custom, natural language chatbots via an intuitive, low-code visual interface for chatbot-enabled insights and conversational user experiences.
Data visualization
Visually explore data, and create and share smart visualizations and interactive reports through a single, self-service interface. Augmented analytics and advanced capabilities accelerate insights and help you uncover stories hidden in your data.
Centralized & searchable model repository
Easily manage analytical models via a centralized, secure web-based repository. Prebuilt model life cycle templates let you manage projects collaboratively.
Task automation with custom workflows
Define and track custom workflows for all phases of model life cycle management – from problem-statement creation to model development and utilization.
Build once, deploy everywhere
Easily deploy models into business processes in a few clicks with rapid, automated model deployment – in batch or real time, in the cloud or at the edge.
Programming-only interaction through REST APIs
Use REST APIs to access, compare, assess and score models.
Cloud native
SAS Viya's architecture is compact, cloud native and fast. Whether you prefer to use the SAS Cloud or a public or private cloud provider, you'll be able to make the most of your cloud investment.
Customer Success
Look Who's Working Smarter With SAS Model Manager
-
"With SAS, we’re working smarter – we’re seeing things that exist in our information that we couldn’t find before, so we can do things more efficiently and effectively, and drive better results for our customers."
David Pardue, Vice President of Connected Vehicle & Uptime Services, Mack Trucks
-
"SAS is adaptable through the choices it affords in techniques, data sources, deployment and even programming languages, while also delivering the speed and scalability that allows us to control outcomes."
Paul Reed, Principal Technical Manager, USG
-
"Because SAS makes it easy for people to do their own modeling, it’s translated into millions of dollars in value for us."
Tim Berryman, Head of Decision Analytics, Georgia-Pacific
Explore More on SAS Model Manager & Beyond
E-Book
Getting Started With ModelOps
Learn how taking a ModelOps approach to deployment moves quality analytic models through development, validation, deployment and monitoring as quickly as possible.
Blog
What's New in SAS Model Manager
Find out about new features and capabilities of SAS Model Manager in SAS Communities.
Webinar
Model Deployment & Management: What You Didn’t Learn About Machine Learning in School
Learn how to get the most value from machine learning with automated, collaborative model management and governance.
E-Book
Mastering the Model Life Cycle Orchestration Journey
Learn about best practices for machine learning in production.