Congratulations to the AI Wonder Girls for winning first place in the #MLOPsforGood Hackathon!

Announcing the Winners of the MLOps for Good Hackathon

Sahar Dolev-Blitental | July 21, 2021

We just wrapped up the first-ever MLOps for Good hackathon, and we are so thrilled by the incredible response we’ve gotten from the ML community. 

300 participants from all parts of the globe—From New Zealand, USA, Canada, Germany, Singapore, India, Portugal, Philippines, Malaysia, Australia, Morocco and Pakistan—joined our cause of bringing data science to production for social Good. After 6 weeks of intense hacking, 30 innovative projects were submitted. It is gratifying to pause and celebrate the ingenuity these hackathoners have shown.

Last night we held the virtual Finals and Awards Ceremony. At the ceremony, our global audience heard from our panel of expert judges, including:

Cecile Blilious - Head of Impact and Sustainability at Pitango Venture Capital

Greg Hayes – Data Science Director, Ecolab

Orit Nissan-Messing - Co-Founder and Chief Architect, Iguazio

Nick Brown – Senior Data Scientist, IHS Markit

Anna Anisin - Founder, DataScience.Salon

Yaron Haviv – Co-Founder and CTO, Iguazio

Adi Hirschtein - VP Product, Iguazio

They also heard from our keynote speakers, Tomer Simon, Chief Scientist at Microsoft Israel R&D Center, and Boris Bialik, Global Head of Enterprise Modernisation at MongoDB. Simon leads the Microsoft AI for Good program, and shared how Microsoft is donating resources and technology to solve global challenges relating to the environment, accessibility, healthcare, cultural heritage, and humanitarian issues. The goal is to “amplify human ingenuity, and bring it to scale [with AI]” he said. He encouraged the hackathon participants to check out the grants available with the AI for Good accelerator program, which kicked off their second cohort in April 2021. 

Bialik shared how easily ML teams can build data platforms in multiple environments, and MongoDB’s efforts to make data accessible for social causes. The most famous example is perhaps the Johns Hopkins University COVID-19 data, which has been made available on MongoDB Atlas for anyone to make use of. Bialik also shared the MongoDB university, with resources that can help anyone gain more proficiency, whether they’re starting from scratch or have lots experience. MongoDB is very active in supporting the next generation of founders, with free access to lots of tools to get started, advice from MongoDB experts like Bialik, and the ability to tap into their mature developer ecosystem.

Finally, we announced the hackathon winners!

The Winners

In first place we have the AI Wonder Girls, with their project to alleviate ICU overload. Their approach to this challenge is to automate the analysis of patient data sets so that healthcare workers in busy ICUs (especially those overloaded with COVID-19 cases) can make quicker decisions and triage urgent care. 

The project had two parts:

  1. A research study analysis model that uses NLP to analyze the associations between Diabetes Mellitus and COVID-19, and highlight factors related to these conditions. 
  2. A model that predicts whether a patient has Diabetes Mellitus based on the patient’s data collected from the hospital admission process. This chronic disease is a co-factor with acute COVID-19 cases, so directing these patients for appropriate care is important for ICUs struggling to keep up with high patient volume. With faster diagnoses, clinicians can plan for better treatment and reduce mortality rates. 

Second place was won by Suicidal post detection, a project that analyzes text to detect whether it contains a suicidal thought, and its severity. The team created an open API that can be used by anyone. This project used MLRun to automate the entire ML pipeline, and uses two different models for training, selecting the best model for serving, and even includes a step for retraining once a certain amount of data has been collected.

Third place was won by Deepfake Shield, with a tool that uses deep learning to detect whether an image contains a deepfake. The team built a pipeline with two models, a pre-trained BlazeFace model (which can be retrained if needed) that extracts faces from images, and a customized implementation of EfficientNet that classifies the extracted faces accordingly. Try out the web app here.

We’d like to give a big thanks to our wonderful partners Microsoft and MongoDB, the 12 incredible communities that joined our cause – TMLS, ODSC, Data Science Salon, Data Talks Club, AI Camp Community, AIIA, MDLI, JerusML, Kinneret Innovation Center, Data Science Initiative, SF Big Analytics & Data for Good, our expert panel of judges, our hard working participants, the Iguazio team and everyone working hard behind the scenes to make this such a success.

We hope to see you in our community and hope that this will be just the start of a movement for MLOps for Good!