Unlock Non-Public Data for research with OpenMined and NAIRR

Gain access to a 1000X more data for your research by joining the National Artificial Intelligence Research Resource (NAIRR) pilot in partnership with OpenMined. You will benefit from access to our flagship software, PySyft, combined with computing credits, comprehensive training and support.

Who is OpenMined?

OpenMined is a fully distributed, open-source community of over 16,000 engineers, researchers, community organizers, product managers, writers, teachers, and enthusiasts. We are actively creating open-source software that uses privacy-enhancing technologies (PETs) to facilitate external access and governance of non-public datasets and AI models, allowing such to be networked together and collectively utilized in transformative ways.

We believe that lowering the barriers to remote data science through our software will lead to a 1000x increase in the amount of data available across scientific fields, public services, and industrial applications, driving a decade of breakthroughs across all of science and industry.

How does our software work?

OpenMined's Expertise

Through our partnership with the NAIRR, we are offering infrastructure and education services for small or large consortias (at a minimum, groups of two or more researchers) to use each others’ nonpublic data for research. Through our collaboration, you will receive:

  • A slice of compute credits for each of your members (if needed).
  • Software to facilitate distributed, privacy preserving data science across the consortia’s nonpublic data (PySyft).
  • A training session teaching your members how to perform distributed learning and privacy preserving data science across the data hosted by your consortia’s members.

Recommended Use

Our resources are best suited for pairs (or more) of researchers who would like to jointly perform research while keeping each others’ data nonpublic. Example use cases may include:

  • Joint Model Training: Consortia of academics can train models on data distributed across various institutions.
  • Research Reproducibility: Academics can allow peer researchers to study their data without direct access, enhancing research reproducibility and integrity.
  • Model Evaluation with Private Datasets: Researchers who would like to evaluate a model while keeping the evaluation data set private.

How to Access Our Resources

If you're part of a research consortium looking to leverage private data for AI research, we invite you to apply for our collaboration credits. Here's how you can get started:

 

  1. Fill Out the NSF Application Form: Provide details about your research consortium including the partners and non-public datasets you would like to work with
  2. Access Resources: Once approved, you'll gain access to OpenMined resources, including PySyft, compute credits, and specialized training.

If in doubt, you can reach out to our team by posting a message in #nairr-resource-support channel on slack to discuss your needs and how OpenMined can best support your project.