MLOps Live

Join our webinar on Implementing a GenAI Smart Call Center Analysis App - Tuesday 27th of February 2024 - 9am PST / 12 noon EST / 6pm CET


MLRun is the first end-to-end open source MLOps orchestration framework. MLRun offers an integrative approach to manage your machine-learning pipelines from early development to deployment to management in your production environment. It offers a convenient abstraction layer to a wide variety of technology stacks and empowers Data Engineers and Data Scientists to define the features and models, simplifying and accelerating the path to production.

Github   |    Join the Slack Community

Open-Source MLRun
Faster Development to Production Through Automated and Scalable MLOps Orchestration

open source orchestrationAnalytics, AI, ML, MLops

  • Central Management: provides a unified portal for managing the entire MLOps workflow. The portal includes a UI, a CLI, and an SDK, which are accessible from anywhere.
  • ML Pipeline Automation: automates data preparation, model training and testing, deployment of real-time production pipelines, and end-to-end monitoring.
  • Elastic Serverless Runtimes: converts simple code to scalable and managed microservices with workload-specific runtime engines (such as Kubernetes jobs, Nuclio, Dask, Spark, and Horovod).
  • Feature Store: handles the ingestion, processing, metadata, and storage of data and features across multiple repositories and technologies.

Why MLRun

Cut Time to Production

Cut Time to Production

Automate MLOps

Automate MLOps

Track and Reproduce

Track and Reproduce

Collaborate and Re-Use

Collaborate and Re-Use

Managed MLRun on Iguazio
Resiliency, Security and Functionality for the Enterprise

Managed MLRun on Iguazio

Enterprise Management & Support

Operators can painlessly set up the system through wizards, configure administration policies and register for system notifications, with no need for automation scripts or hands-on daily management. The Iguazio Data Science Platform with managed MLRun is delivered as an integrated offering with enterprise resiliency and functionality in mind. Enable data collaboration and governance across apps and business units without compromising security or performance. Authenticate and authorize users with LDAP integration and secure collaboration. The real-time data layer classifies data transactions with a built-in, data firewall that provides fine-grained policies to control access, service levels, multi-tenancy and data life cycles. Enterprise customers get dedicated 24/7 support to onboard, guide, and consult.

Data Layer 

Achieve extreme performance with consistency at the lowest cost with the real-time data layer, built-in to enterprise-grade MLRun. With key-value and time series databases and an object store available out of the box, the data layer supports simultaneous, consistent and high-performance access through multiple industry standard APIs. The data layer provides fast, secure and shared access to real-time and historical data including NoSQL, SQL, time series and files. It runs as fast as in-memory databases on Flash memory, enabling lower costs and higher density.

Managed Services

Lift the weight of infrastructure management by leveraging built-in managed services for data analysis, ML/AI frameworks, development tools, dashboards, security and auth services, and logging. With managed MLRun, anyone on your ML team can simply choose a service, specify params and click deploy. Data scientists can work from Jupyter Notebooks or any other IDE and automatically turn it into an elastic and fully managed service directly from Jupyter or another IDE, with a single line of code.

MLRun Ecosystem

Open Source VS Enterprise

Features & Functionality
Open Source MLRun
Managed MLRun on Iguazio

Project Management

Single pane of glass for managing projects
Git integration
Built-in services (Spark, Presto, Grafana, Dask, Horovod etc.)
Project resource management
Air-gapped environments support
Built-in Jupyter service

Experiment Tracking

Artifacts management
Pipelines management

Model Deployment

Serverless function for running models (nuclio)

Model Monitoring

Model monitoring dashboard
Model drift identification
Canary rollout

Feature Store

Offline features for training
Online features for serving

Built-in Data Layer

Key-value database
Time series database
Built-in object store


LDAP integration
User(s) & group management
Service authentication & authorization
Secured authentication for API gateway

Management & Support

Community support
24/7 enterprise support
Service monitoring
Logs management
Performance monitoring reports
Events, alerts & audit

MLRun Resources