Hello fellow hunters! As a machine learning scientist, I’ve learned that there is never enough data. Machine learning and analytics have so much promise to solve some of the world’s biggest problems - from precision medicine, to drug discovery, to energy efficiency. However, privacy, confidentiality, and technical hurdles make it hard to access all this data for training. We’ve some of these barriers with data sharing agreements, but they’re complex, hard to enforce, and slow projects down to a standstill. In response to this, we’ve built a privacy-preserving machine learning and analytics platform to help overcome the hurdles of using distributed data that is hard or impossible to access. With integrate.ai: - Training is performed at the data nodes so raw data is always under the control of the data custodian - Differential privacy is added during training to further protect the final model from deidentification - Upload your own model for training and we will get it federated - Import our SDK and use your existing workflow We’ve been working on this for over a year now, and we’re excited to share what we’ve built with this community! Thanks for all the support! And we’d love to answer any questions you may have! Ro, VP of Product @ integrate.ai

Hello Hunters! I'm Steve, the founder and CEO of integrate.ai. I wanted to share a bit about our mission and where this product launch fits in, so you know what you are supporting when you upvote us or comment below. At integrate.ai, our mission is to accelerate the shift to AI in a way that positively changes the lives of more people. We have a goal of changing 1 million lives in the next 1,000 days by enabling our ecosystem of developers to bring innovation to critical industries and use cases. We have talked to a lot of developers working on some of the world's most important (and difficult) use cases from rare disease diagnosis, to personalized autism treatment, to early detection of mental health crises, to alternative credit scoring to bring more unbanked into the financial system, to lower carbon emissions in industry, to safer equipment in factories, and more - and all have the same challenge. They need access to more sensitive data and collaborations to see the full potential of AI for those use cases. Often this data is trapped in multiple silos (across jurisdictions, or partners, or hospitals, or IOT devices) and can't be centralized due to privacy, security, or IP concerns. The integrate.ai platform allows developers to run machine learning or analytics solutions across data that is distributed across many silos without needing to move the data. This unlocks these meaningful use cases that can directly save, extend, and enhance people's lives. By helping get us more visibility, we can onboard more developers working on these important use cases to our platform and ultimately see more impact in the lives of people we know and love. If you are one of those developers, we would love to support you in bringing innovation to these important areas. If you know someone working in an important area where data is sensitive or hard to access, but AI or advanced analytics could meaningfully accelerate their solution - please comment below and tag them. We would love to talk to them. Thanks for your support. We appreciate everyone that showed up to help us today!

Roshanak Houmanfar

integrate.ai

Maker

3yr ago

Steve Irvine

Vlad Nitu

Hello Hunters! It is often said that "Data is the new oil". However, if unrefined it cannot really be used... Many hospitals and medical centres generate large amounts of data which, if refined, have the potential to cure diseases and lead to better medicines. However, because of the strict regulations, this data is often kept "under the ground" The privacy-preserving ML platform built by Integrate.ai is able to refine this data while complying with the healthcare regulations because data never leaves the hospital where it is generated. This is a good example on how AI can make the world a better place, let's support it!

integrate.ai

Extend your product to train ML models on distributed data

Do you use integrate.ai?

Review integrate.ai?

Reviews of integrate.ai