As we face the most severe global health crisis many of us will see in our lifetime, we recognize the role AI can play in addressing challenges posed by COVID-19. We invite data scientists, machine learning practitioners, developers, health IT professionals, public health experts and others around the globe, to participate in the Lumiata COVID-19 Global AI Hackathon.
Participants will develop solutions that address the challenges this pandemic poses to citizens of the world today and for the impact it will have in the future. Winners will be recognized with solutions fully open-sourced and hosted on on the Lumiata AI Platform and published with full credit to the developer, as well as public recognition in an awards ceremony.
We encourage participants to develop AI solutions for both current challenges posed by COVID-19 as well as longer-term concerns that will result following the immediate effects of the crisis. The challenge is to dream big and focus on tangible solutions that impact population health as well as business challenges.
Examples include but are not limited to:
• What resources can we provide to health systems (providers, insurance companies, etc) and their staff to prepare them to care for patients with COVID-19 in the next 1-3 months? 3-6 months? 6-12 months? Beyond?
• How do we establish an intelligent credibility score for online COVID-19 information offered to the public? Imagine - You submit a URL with a COVID-19 article, and it generates a credibility score.
• How will deferred elective procedures impact the health system and health plans once those sheltering at home re-engage with the health system?
• What is the projected financial impact of COVID-19 on health system stakeholders such as Medicare, Medicaid, small hospital systems and self-insured employers?
• How will the economic shut down affect the health system?
• Can we identify any early predictors of COVID-19 as indicators of future pandemics?
• How can governmental leaders measure the impact of decisions like lockdowns and social distancing
• How can data scientists, analysts, and actuaries create models with an incomplete picture from a data perspective, how can they model with data that does not exist yet?
• Can we predict medical staff burn-out and mental health issues due to the crisis?
• How can we speed up the clinical trial process in general? and more specifically, the recruitment process?
• How can AI / ML help researchers working on disease investigation?
These are some of the questions that we are hearing from customers. Do not let the specific questions limit your imagination, please suggest new ideas thinking about health system stakeholders.
Heath System Help
Not everyone will have health system knowledge nor expertise, this is why we will have health system experts answer questions you may have in the Slack channels.
You can find links to open data sources for the Hackathon in our Lumiata Hackathon page:
Here is a consolidated list of data sources and resources:
$12,500 in prizes
First Prize, $10,000 and idea productionalized
We will help the winner bring the idea to production with a real world Healthcare / Public Sector customer or partner.
Runner up, $2,500 and idea productionalized
We will hand the cash price, and we will help the runner up bring the idea to production with a real world Healthcare / Public Sector customer or partner.
Submitting to this hackathon could earn you:
- Participants: Individuals (over 18 years of age, or over 13 with a legal guardian's consent), and Teams of up to 6 people (Pizza size teams).
- Teams: You can bring your existing team to the Hackathon, or you can reach out to other participants to form a team. Whether you're a data scientist and need help with software engineering or a software engineer and need machine learning expertise, you can build a team or we can help you identify partners.
- Countries: Worldwide.
This hackathon welcomes brand new ideas that will be built during the hackathon time-frame, as well as projects that individuals or teams may already have in-flight.
We expect that submissions are proof-of-concept quality software and must showcase the core idea.
The following are the supported submission types:
- Working software, an end-to-end working piece of software that can be executed in a target machine, it can have a command line or a GUI interface
- Machine learning model(s) tackling a specific problem (we encourage novel solutions and algorithms beyond standard ML and Neural Networks models)
- A report or data visualization that illustrates some data analysis that addresses a specific problem
The following will be required for every submission:
- A write-up describing the submission
- A link to a public GitHub repo with code for the submission
- Instructions on how to run the project
- A short video describing the idea and why it makes a difference
Founder, Khosla Ventures
President, Care Journey Former Chief Technology Officer, United States
Professor of Mathematics, University of California, Berkeley Member of the American Academy of Arts and Sciences Author, Love and Math
Edward H. Bowman Professor Professor of Management Vice-Dean, Wharton Social Impact Initiative
Severence M. MacLaughlin, Ph.D
Managing Director, DeLorean Artificial Intelligence
Lead, Digital & Technology Group Lead, Manatt Ventures Managing Director, Manatt Health
General Manager, Intel Health & Life Sciences
Chairman & CEO, HMS
Founder and CTO, Question.ai
Global Director, Healthcare Solutions, Google Cloud
Chief Information Officer, California Health and Human Services Agency
How unique is the submission? Is it really something unique, and new? How much is it pushing the boundaries of the status quo? How much are you pushing your imagination?
It is required that all submissions have some sort of AI orientation. It must have Machine Learning built in part of the core. How strong is that core? For ML models, evaluation metrics will be considered.
How much impact the submission will have once in production. It can be measured in terms of people reach, people helped, money saved, etc. This is a very important item in the criteria list, the best submissions will the ones with highest impact.
Can the submission scale in a real world context? Scale can be measured by number of users for a submission, the number of applications that an ML model can power, number of use cases fulfilled by the submission, etc.
How feasible is it to bring the submission live to production? Feasibility from a technical perspective but also from a context perspective, ie: how feasible would be for a healthcare company to adopt give then context?