New Features for Leading Model Operations System Further Automate and Accelerate Deployment of Analytic Models into Production
Chicago, IL. – June 10, 2019 – Open Data Group, the leader in model operations (ModelOps) systems for large enterprises, today announced the newest version of the FastScore Model Operations System with Release 1.10. The release further enables data scientists, data engineers, and IT operations teams to collaborate and scale analytics across the organization and thereby drive business value by getting models into production faster and with greater visibility, accountability and control.
The FastScore System is the industry’s only platform-agnostic model operations system. It can be used to deploy into production models created in any language, on any workbench or toolkit, using any IT orchestration system. It also provides a full suite of modules for controlling, monitoring, managing, and reporting on models in all stages from development through testing, staging and production. By simply implementing FastScore’s lightweight abstractions, scientists can create models that can be deployed and maintained in production by the IT team, anywhere, without the need for refactoring. This unique, patent-pending approach provides complete freedom for data scientists to use the latest and most appropriate data science tools, and frees the IT operations team to deploy into their existing or new infrastructure – while providing business owners with the visibility they need to track the performance and contribution of analytic assets.
New capabilities in Release 1.10 further accelerates the deployment process by streamlining model on-boarding into FastScore. Through FastScore’s powerful but lightweight abstractions, data scientists can encapsulate their models for easy deployment through the FastScore System without any re-coding or refactoring. Release 1.10 streamlines conformance to FastScore’s abstractions for a large number of additional classes of models. Additionally, Release 1.10 adds automated schema inference to allow data scientists to quickly generate the externalized schemas to ensure compatibility of data streams with the model.
Key benefits of Release 1.10 include:
- Accelerates on-boarding of models into FastScore by providing support for conforming additional classes of models to FastScore’s core abstractions.
- Automates model schema creation allowing users to more rapidly realize the benefits of encapsulation, streamlined data engineering integration, and execution without refactoring.
- Expands support for Pandas by providing native support for Pandas DataFrames as input and output from the FastScore Engine, allowing data scientists to more easily integrate their models into FastScore.
- Adds support for a Go CLI and SDK allowing data scientists and operations teams to leverage the rapidly growing and powerful Go language.
- Enhances the core logging framework to leverage the ELK (Elasticsearch, Logstash, Kibana) stack, providing system architects and support engineers more flexibility to integrate FastScore’s logging capabilities into existing enterprise systems and standards.
“Data scientists want and need the freedom to develop the best analytic models, in their chosen languages and platforms, and to see them deployed quickly and successfully into production at scale, without proprietary platform constraints and without the need for time consuming, error-prone re-coding or refactoring,” said Matthew Mahowald, Director of Data Science at Open Data Group. Matt continued, “FastScore Release 1.10 further delivers on our vision of a world in which analytic models move from development to production at maximum speed and deliver maximum value to the business.”
FastScore Release 1.10 is available starting June 10th. All existing Customers are eligible for Release 1.10 as part of their paid subscription.
- Complete FastScore 1.10 Documentation can be seen here
- Infer Schemas video can be seen here
- Explicit Model Conformance video can be seen here
FastScore gives the enterprise a modern, microservices based approach for machine learning and AI model operationalization. FastScore is architected as a suite of microservice modules based on Docker. Each is an optional, but powerful, service used to connect the critical pieces of the analytics workflow: data science models, data sources, and applications. The underlying philosophy of FastScore is integration: provide unique value where appropriate, and leverage existing technology when available. Read on to learn more about each module, and the value they bring to operationalizing models.