AI-Powered Covid Assessment

The Team

Myriam Bocobza - Team Lead

Ehren Goossens - Team Coordinator

Noemie Ifrah - Research / Strategy

Sarel Cohen - Software Developer / Research Engineer

Yonatan Cale - Software Developer / Data Science

Jeremie Brabet-Adonajlo - Research / Strategy

Helit Rozen - Molecular Biology

Inbal Lidan - Software Engineer / Data Science

Shimon Aronhime - MD

Michael Fiszer - Computer Science / Growth Hacking

Kosta Maslev - Software Engineer

Leah Orlin - Software Developer / Data Science

Shimon Podval - Electrical Engineer

Alexandre Sagal - Research / Communication / Marketing

The Challange

59% of Covid cases in Wuhan were undetected. “Several recent reports also highlighted the difficulty to detect Covid-19 cases: about two thirds of the cases exported from mainland China remained undetected worldwide, and the detection capacity varied from 11% in low surveillance countries to 40% in high surveillance countries.” (SOURCE:  Evolving Epidemiology & Impact of Non-pharmaceutical Interventions on the Outbreak of Covid-19 in Wuhan, China)


"It is abundantly evident now, especially in the sampling of icelandic genetic analysis, that about half of those who were positive were left with little or no symptoms.”  - Pr. Thorolfur Gudnason, Iceland Chief Epidemiologist


We need a better way to identify clusters of mild, asymptomatic and presymptomatic cases, because they are driving the spread of the virus.

Current status

We are finalizing the contents of the simple and in-depth assessments. Contents and Wireframes for the simple version of the assessment (for Hamagen/CoronaIsrael/Klalit) are ready. Contents and wireframes for the in-depth version will be ready 04.19.20. 


Software Devs + Data Scientists are ready to commence ML portion, standing by for data from Weizmann Institute. 

The Solution

We are developing a Covid self-assessment powered by a machine-learning algorithm. Our objective is to analyze user input and predict the likelihood of positive covid diagnosis. We believe by layering differentiated symptom data with other key risk factors and navigation data, we can deliver an accuracy performance equal or superior to current physiological tests - while vastly more scalable and cost-efficient. We are collaborating with the Prime Minister’s Office, the team that built the Hamagen app, and Eran Segal’s team at the Weizmann Institute, who developed the questionnaire for and Klalit. This version of the assessment will be relatively minimal to optimize use engagement. 


We are also building a very detailed Covid assessment to be pushed globally, as both the PMO and Weizmann Institute agree there is an appetite for a more in-depth/lengthy assessment that would enable strategic data collection. 

What’s Next

To get access to Weizmann Institute data and start work on the ML aspect of our project, we need Helsinki Approval. We are finalizing our application. Dr. Tal Patalon has agreed to stand as our Primary Medical Investigator.