Coda’s AI at Work Challenge
Posted: Mon Feb 10, 2025 4:54 am
Participants were asked to build and train an ML model with Determined AI, which is an open-source platform for building and training ML models. The event was open to everyone from experienced ML engineers to those who are brand new to ML, with many resources provided to participants from Determined. Participants were tasked with building a machine learning project (complete with a deep learning-based model, training loop, evaluation loop, and dataset) that works on the Determined AI platform.
The Results: This hackathon kicked off during the beginning of the “AI boom,” so it’s no surprise it garnered a lot of attention. There were 820 participants and three winners. The top prize went to a project that created a model to predict a Parkinson's patient's future UPDRS-3 score, which measures the patient’s motor skills. The macedonia telegram screening model was built using Determined AI’s Cloud service and Tensorflow Keras libraries. The model was trained using the AMP®-Parkinson's Disease Progression Prediction dataset from Kaggle.
See all the projects here.
Company: Coda
Summary: Submit templates that show how Coda AI helps you save time and accomplish more—for a chance at a $10k grand prize.
Coda is an all-in-one doc that brings words, data, and teams together. In this hackathon, participants were asked to submit Coda AI-enabled templates that help people save time and accomplish more. The hackathon allowed participants to experiment with using Coda and find ways to improve it. The templates were required to fit into specific categories and had to be submitted to the Coda Gallery.
The Results: This hackathon had 1036 participants and 168 project submissions that provided new ways to use Coda AI-enabled templates. Overall, there were 10 winners who showcased creative use cases for Coda AI, including marketing, health and wellness, and recruitment. The top prize went to a project that made it quick and easy to extract social post content from long-form content and turn it into ready-to-post Instagram carousel images.
The Results: This hackathon kicked off during the beginning of the “AI boom,” so it’s no surprise it garnered a lot of attention. There were 820 participants and three winners. The top prize went to a project that created a model to predict a Parkinson's patient's future UPDRS-3 score, which measures the patient’s motor skills. The macedonia telegram screening model was built using Determined AI’s Cloud service and Tensorflow Keras libraries. The model was trained using the AMP®-Parkinson's Disease Progression Prediction dataset from Kaggle.
See all the projects here.
Company: Coda
Summary: Submit templates that show how Coda AI helps you save time and accomplish more—for a chance at a $10k grand prize.
Coda is an all-in-one doc that brings words, data, and teams together. In this hackathon, participants were asked to submit Coda AI-enabled templates that help people save time and accomplish more. The hackathon allowed participants to experiment with using Coda and find ways to improve it. The templates were required to fit into specific categories and had to be submitted to the Coda Gallery.
The Results: This hackathon had 1036 participants and 168 project submissions that provided new ways to use Coda AI-enabled templates. Overall, there were 10 winners who showcased creative use cases for Coda AI, including marketing, health and wellness, and recruitment. The top prize went to a project that made it quick and easy to extract social post content from long-form content and turn it into ready-to-post Instagram carousel images.