Description
Cortex is open source infrastructure for deploying and scaling machine learning models in production. It runs on top of Amazon EKS and deploys clusters into a VPC on your own AWS account, so your data stays in your account. You can run models as realtime services that scale with in-flight requests, as async workloads that scale with the request queue, or as batch jobs that run distributed and fault-tolerant processing. Cluster management handles autoscaling with CPU and GPU instances, spot instances with automated on-demand backups, and separate environments for different configurations. You can provision clusters with declarative configuration or a Terraform provider. Metrics can go to any monitoring tool or to pre-built Grafana dashboards, and logs can be streamed to any log tool or to CloudWatch. It is aimed at teams that serve machine learning models, retrain them, or process large batch jobs and need them to scale reliably.…
Highlights
- Realtime, async and batch workloads
- Autoscaling with CPU and GPU instances
- Runs on AWS EKS in your own VPC
- Terraform provisioning and Grafana dashboards
- Model serving and continuous retraining
Ratings & Reviews
No ratings yet
Be the first to rate Cortex.
People also visit
Datatron
Build, deploy, automate ML models to identify patterns, make predictions
Cortex FAQ
What is Cortex?
Infrastructure for deploying, managing and scaling machine learning models in production on your AWS account.
Who is it for?
Teams that serve machine learning models or run batch processing and need it to scale reliably.
Does it run on AWS?
Yes. The site says it runs on top of Amazon EKS and can deploy clusters into a VPC in your AWS account.
Is it open source?
The site links to a public GitHub repository for the project.
What are the best Cortex alternatives?
Popular alternatives to Cortex include AI Surge Cloud, Datatron and Metaflow. See all Cortex alternatives









.png)
