Open Nav

Why we decided to release a free ML platform to the data science community

Why we decided to release a free ML platform to the data science community

When we started building a little over 3 years ago, Leah, my co-founder and I, have always dreamt of releasing a version of to the public, for free. As developers and data scientists ourselves, our mission with has always been to help data scientists and data engineers apply their research and innovation to real world applications. Today, we’re extremely excited to announce the release of core – a free community ML platform.

We started to build as an internal solution to help manage our own models, research, and machine learning in production. We saw how beneficial it was to our data science workflow, and believed in its potential to transform the data science development as we know it. And so began our journey, working with early customers to provide everything a data scientist needs to accelerate ML development from research to production.

Over the past several years, we’ve been doing just that. Working on building, productionizing and improving based on customers demand and market needs. today is different from the we had 3 years ago, 1 year ago, and even 3 months ago. We’re constantly working to improve our services to meet the needs of data scientists. This crazy, fast-paced journey we’re leading as a product and as a company is correlated strongly with how AI is evolving in the enterprise, and how data scientists and engineers build and deploy models.

Our dream is to help any data science team to build, deploy and manage machine learning – from research to production. With the growing technical complexity of the AI field, the data science community has strayed from the core of what makes data science such a captivating profession – solving complex problems with ML. From our experience working with DS teams across industries is that today’s reality is data scientists are spending 80% of their time on non-data science tasks, and 65% of models don’t make it to production. We want to make data science workflows extremely simple and intuitive so teams are able to focus on innovation. We want to create an infrastructure with best-practices in mind of how models, and Software 2.0 should be built, so teams can release high impact models to production. We’re extatic to share the releasing of CORE – a free, community version of that can be used for any use case. Whether you’re a team of data scientists building autonomous driving applications, churn-prediction models, forecasting or natural-language and image processing your team can utilize all the features of CORE.

Now more than ever, we’re in a time that requires innovation from our data science leaders. The world is seeking solutions as we reach more and more global crises. Our vision is to give data scientists all the tools they need to collaborate and do what data scientists do best – solve complex problems with machine learning. While the rest of the world has slowed down, we hope innovation can accelerate. We welcome any data scientist to install CORE in any environment, whether working remotely or not. The platform is created with flexibility in mind, and can be easily installed on any Kubernetes cluster – cloud or on-premise. This version of includes everything a data science / data engineering team needs to build and manage machine learning pipelines and models. With, a container-based ML platform with end-to-end monitoring, model management and advanced resource management, your team can focus on what is truly important – innovation. While has everything a data scientist needs to build high impact models, with cnvrg you can integrate to AWS Sagemaker, MLFlow or any other ML stack to enhance your existing ML workflow. Any team can adopt our unified, collaborative platform to accelerate time from research to production – no matter where you’re working from.

What is CORE?

The CORE platform offers a unified code-first workbench built for data scientists. Supplied with workflow management tools, pre-configured containers and cluster orchestration, data scientists can begin exploration quickly. CORE’s flexible infrastructure gives data scientists the control to use any language, AI framework, and compute environment whether on premise or cloud compute. CORE offers everything a data science needs to build high impact ML models from research to production with:

  • ML Workflow Management – end-to-end tracking and monitoring capabilities
  • Interactive Workspaces – use any development environment like JupyterLab, RStudio, and more with pre-installed dependencies and version control
  • Dataset Management – integrate and version any kind of data to reuse in any project, experiment, and/or notebook
  • ML Flows – build production-ready machine learning pipelines with drag and drop
  • AI Frameworks in One-Click – native integration with NGC GPU-optimized containers 
  • Reusable ML Components –  Build and customize a library of reusable ML components 
  • Cluster Orchestration – Hybrid cloud / Multi Cloud capabilities with native Kubernetes and meta-scheduler  
  • One click Deployment – deploy models as  scalable REST APIs based on kubernetes  
  • Production Model Monitoring – monitor and manage production models with Kibana, Grafana and update with continual learning
  • Launch 100s of experiments – track and compare hundreds of experiments with hyperparameter optimization

    After years of building and growing, we’re proud to unleash its value to the strong data science community that’s been driving AI innovation and solving complex problems. Our hope is that the field of AI continues to grow. will do everything possible to support its growth, by supporting the data scientists and engineers behind its success. 

Yochay & Leah

Top MLOps guides and news in your inbox every month