Comet Expands Alliances Across the Complete Machine Learning Lifecycle

New integrations with Ray, Kubeflow, and Google Vertex AI strengthen Comet as the only combined experiment management and model production monitoring solution with true enterprise scalability and extensibility

NEW YORK--()--Comet, provider of the leading development platform for machine learning (ML) teams from startup to enterprise, today announced several integrations – including Ray, Kubeflow, and Google Vertex AI – that will eliminate complexity, provide optionality, and bolster scalability across the entire machine learning lifecycle.

These integrations deliver on Comet’s commitment to providing a ML platform that allows data scientists and ML practitioners to use the best tools and frameworks available for their use cases, industries, and business challenges from experimentation through production.

Comet’s unique approach to integration comes at a time when more organizations are pursuing the promise of artificial intelligence (AI) while their development teams are increasingly encountering friction between tools and among teams. This friction is the result of a rapidly evolving MLOps technology landscape and the lack of processes that support the burgeoning needs of companies looking to generate value from AI.

Too often, teams in fast-growing organizations become siloed, and their use of tools and processes become even more disconnected and inefficient. Comet addresses this issue through strategic integrations, making it seamless for teams to utilize the tools that best fit their specific business requirements, integrated within a single platform. This approach reduces the friction, allowing teams to reduce the time to achieve value from their AI initiatives, and enables new levels of management and insight, fostering better collaboration across teams.

“In this early adoption phase of AI/ML software, we’re seeing a massive proliferation of tools and platforms that were created to address the unique, highly specific issues of data science. While these are effective in solving pieces of the puzzle, there hasn’t been a comprehensive solution. Enterprises are struggling to cobble together the dozens of different tools needed to make a complete ML platform as a result,” said Daniel Jeffries, Managing Director of the AI Infrastructure Alliance (AIIA). “For ML to reach its potential, deeper interoperability between enterprise products is very much needed now. Tools that work together extremely well is the best approach to scaling ML for all.”

The Comet platform makes it possible for ML development teams to track, compare, debug, and monitor models from early experimentation all the way through production. It uniquely offers both standalone experiment tracking and model production monitoring, and its platform can run on any infrastructure, whether it is cloud, on-premises, or virtual private cloud (VPC). The latest integrations join what is already the most extensive list in the industry. In addition to Ray, Kubeflow, and Vertex AI, Comet enables customers to leverage the best frameworks, libraries, and compute resources – such as Amazon Sagemaker, Catalyst, GitLab, Keras, New Relic, PyTorch, Scikit-learn, Spark NLP and Tensorflow.

New Tools, Better Insights in One Comprehensive View

Comet’s most recent integrations provide enterprises with necessary capabilities for developing a cohesive ML platform. For example:

  • Ray: An open source project, Ray makes it simple to scale any compute-intensive Python workload — from deep learning (DL) to production model serving. The Comet integration allows data scientists to leverage Comet’s experiment tracking and visualization tools with Ray’s library for scaling compute intensive ML workloads.
  • Kubeflow: An open source ML platform, Kubeflow enables using ML pipelines to orchestrate complicated workflows running on Kubernetes. The Comet and Kubeflow integration lets users track both individual tasks and the state of the pipeline as a whole.
  • Vertex AI: Google Vertex AI enables users to build, deploy, and scale ML models faster, with pre-trained and custom tooling within a unified artificial intelligence platform. Comet partnered with Vertex AI to allow users to track not just individual training runs but also Vertex Pipelines within the Comet UI.

The integrations with these tools help companies reduce information silos by bringing data from all of the tools in use for ML development together within the Comet ML platform.

“The majority of ML platform teams out there are trying to figure out how to assemble an end-to-end MLOps stack that fits their needs,” said Gideon Mendels, CEO and co-founder of Comet. “With different use cases, dependencies on existing tools, and an increasing number of vendors, that is a very challenging task. Comet is the only modular solution that allows these teams to use their desired tools while still providing a unified user experience to ML practitioners. We do this by providing built-in integrations to support features, like our code panels, that extend practitioners’ visibility across the ML lifecycle.”

To learn more about the Comet platform and how it can help enterprises fully leverage AI, ML, and DL, visit www.comet.ml.

About Comet

Comet provides an MLOps platform that data scientists and machine learning teams use to manage, optimize, and accelerate the development process across the entire ML lifecycle, from training runs to monitoring models in production. Comet’s platform is trusted by over 150 enterprise customers including Affirm, Cepsa, Etsy, Uber and Zappos. Individuals and academic teams use Comet’s platform to advance research in their fields of study. Founded in 2017, Comet is headquartered in New York, NY with a remote workforce in nine countries on four continents. Comet is free to individuals and academic teams. Startup, team, and enterprise licensing is also available. To learn more, visit www.comet.ml or join our community at heartbeat.comet.ml.

Contacts

For Editorial Contact:
Amber Moore
GMK Communications for Comet
amber@gmkcommunications.com