Blog

doc.ai Launches Crestle.ai, a One-Click Deep Learning Platform

Doc.ai runs data trials for AI-powered medical research, using a hosted analytics platform for data scientists. To enhance this platform, doc.ai has acquired Crestle (now Crestle.ai), a popular data science analysis platform, and has rewritten it for greater scalability to ensure it meets the needs of the doc.ai data science community. Crestle has 11,000 registered data scientists, with a user growth rate of 10% per month.

This platform can support not only medical AI research, but a variety of use cases. Therefore, we are making it available to the global data science community. Data scientists now have instant “one-click” access to a scalable, powerful, inexpensive GPU and CPU infrastructure. The new platform is already being offered to students of the new fast.ai Live course, which started on October 22, with thousands of students from all over the world. Along with fast.ai Live, Crestle.ai will support new deep learning courses in 2019 for physicians and bioinformaticians, as well as the next fast.ai Practical Deep Learning for Coders MOOC. Crestle.ai previously provided its platform under the domain name crestle.com, which will continue to be supported for general purpose deep learning.

Our consumer facing doc.ai app and its thousands of users including a community of more than 1,800 participants for our first data trial is among our first efforts to bring a synergy between AI professionals and health industry domain experts. This kind of convergence of artificial intelligence and healthcare is inevitable, but we need the domain experts in healthcare to make a major contribution to advancements in AI.

As a step towards this mission, a new series of deep learning, data science, and decision science courses for physicians is to be released in 2019 by Professors Chirag Patel and Arjun Manrai from Harvard University, expert in bioinformatics and advisor at doc.ai. This will make AI and integrative genomics tools accessible to a wide number of people in medicine, an industry where lack of resources remains a core challenge.

According to The New York Times, AI is essentially a community of 10,000 people. It’s even more narrow in terms of domain experts. In most cases, data scientists are coming into the field with a background in software engineering or from academia. They might be skilled and knowledgeable in machine learning, statistics, math/computer science, but in most cases they lack the background in healthcare while physicians lack the background in machine learning, statistics, math/computer science.

The challenge of the healthcare industry with this huge amount of clinical knowledge is that there are only a few people in the world who are able the comprehend the data and professionally work with it using limited medical terminology to get the necessary insights to solve real-world problems.

That’s why it’s crucial for the industry to build an ecosystem where AI intelligence and the healthcare domain meet. With our Crestle.ai infrastructure, we are aiming to create an environment for highly skilled AI physicians through providing educational courses and an easy to use Deep Learning infrastructure.

Launch of Crestle.ai for the Fast.ai course early this week with our CTO Akshay Sharma with our team and doc.ai open source advisor Frederick Kautz.

Crestle.ai provides a familiar Jupyter-based computing platform on a browser. It comes pre-installed with all the popular scientific packages, machine learning libraries and deep learning frameworks. It is ready to go in one click! Additionally, it provisions all the necessary datasets for you to take fast.ai course (many orders of GBs), and you don’t have to pay for the datasets! You also get a scalable home directory to bring additional datasets to build your model, which is reliably backed up for you with zero infrastructure effort on the participant’s side. Doc.ai’ s AI and infrastructure teams have strived to make this a scalable platform for everyone.

A few months ago we shared about our collaboration with Anthem using our AI tools and our blockchain infrastructure, following the course of this collaboration Anthem will also make Crestle.ai available to its community of developers and physicians. “It’s part of Anthem’s efforts on leading our industry in the safe and responsible use of artificial intelligence and emerging technologies to create a better healthcare future for all Americans.” says Rajeev Ronanki, Chief Digital Officer at Anthem.

Crestle.ai’s infrastructure perfectly fits doc.ai’s mission to advance medical research. Our first data trial ‘Can AI Predict Your Risks for Allergies?’ is currently running with the support of our advisors from Harvard University, a framework using AI to predict and track allergy risks of more than 1,800 participants based on work by Chirag J. Patel, Nam Pho, Michael McDuffie, Jeremy Easton-Marks, Cartik Kothari, Isaac S. Kohane & Paul Avillach (Oct 2016), and a database of human exposomes and phenomes from the US National Health and Nutrition Examination Survey (Nature Scientific Data Volume 3, Article number: 160096).

That amount of medical data will help create targeted and advanced educational patient tracks for physicians and offer the community highly skilled AI professionals with a unique domain knowledge. And the blockchain technology on which the data trial is based is ensuring the security of all shared data. We’re looking forward to share more about that part in future posts.

The doc.ai AI powered medical research companion app is available on the app store here. Join thousands of people making the first steps in collecting and owning their medical data to get personalized insights and help accelerate research.

If you’re a data scientist check Crestle.ai the one click deep learning platform, it’s free until next week including free GPU, available here.

Check out the article at VentureBeat!