Welcome to the Nightingale OS Platform

Connecting computational researchers with clinical health care data

Launch (NeurIPS 2021): Machine Learning from Ground Truth

Register Now

When you register, you enter a queue for access to the data. We will grant access to researchers incrementally, starting in late December and picking up speed in January.

Before granting access, we'll ask you to provide the following:

  • Contact information for a person at an academic institution who can verify your non-commercial use case for Nightingale OS /datasets
  • Certificate of completion for a suitable training program in human research subject protections, such as the Collaborative Institutional Training Initiative (CITI) Program's Data or Specimens Only Research course

Our support team will verify your information, including reaching out to your institutional contact. We will admit people as quickly as we can. Please be patient as we ramp up computing capacity to meet user demand.

Work in a familiar environment

We designed the platform with extensive input from ML researchers. If you've used Python in JupyterLab, then you won't need to learn anything new. If you've ever spent weeks setting up a research environment for your lab, then you'll be shocked at how quickly you can get started in Nightingale OS and how easily you can onboard collaborators in your research projects.

Free cpu.xsmall instances

Once you are admitted to the platform, you will immediately get access to our free cpu.xsmall (CPU-only) compute instances with a Python environment in JupyterLab. You can use these instances to explore all of our exciting new datasets and plan your approach.

Larger computing resources coming soon

Beginning in January 2022, you will be able to do more data-intensive tasks using larger CPU-only instances with up to 30 CPUs and 250 GB of memory. You will be able to generate an invoice that you can pay or forward to someone else to fund your projects. Once you've funded a project, you will be able to take advantage of new cpu.small, cpu.medium, cpu.large, and cpu.xlarge instances.

As a nonprofit, we are providing compute at as close to cost as we can. For researchers who need additional help, we are working with our funders to provide scholarships that ensure open, equitable access to use Nightingale data. Please tell us if you need assistance.

GPUs shortly after

We plan to introduce gpu.small instances in roughly late January, although the timeline will depend on our user metrics and GPU sourcing.

You may be dealing with this problem elsewhere, but GPU capacity is in short supply. Our goal is to provide near 100% availability of gpu.small instances a few weeks after they are introduced. In the first few weeks, you should expect GPU instances to be scarce. They'll be first-come first-serve.

As we measure demand across the platform, we will work with our providers to secure appropriate resources. After we achieve near 100% availability of size gpu.small, we plan to introduce instances with more GPUs later.