What to Build

 Build a new Generative AI project with NVIDIA AI Workbench and post it on GitHub. The Workbench Project should leverage AI Workbench’s ability to make it easy to get up and running in generative AI applications and drive developer productivity for data science, AI, and machine learning tasks and workflows. Demonstrate the Workbench project running on a GPU system of your choice.

What to Submit

 Provide a URL to your Project on GitHub for judging and testing. Keep your Workbench Project private on GitHub. You will be sharing it with Devpost and NVIDIA.

• Make sure the Project has an explanatory README with instructions on how to run the Project in AI Workbench. Include a text description that explains the features and functionality of the workflow or application you built in your Project. README should indicate target systems needed to run the Project, as well as any sort of restrictions in terms of models or containers used.
• You can see an example here for a RAG Project, as well as one for a multi-container application here.

 Include a video (about 3 minutes) that explains and demonstrates your project in action, running on a GPU system of your choice. Videos must be uploaded to YouTube or Vimeo and made public. Social media handles for all team members: LinkedIn, Discord, Twitter. Optional -  blog post included in your GitHub repo. It must be 800-1200 words and should address what was created and how it was created.  Your complete submission form on Devpost before the final deadline.  Non-Category Bonus Prizes

• Most Valuable Feedback Prize: If you have technical tips or tricks, or any other feedback, share it in the Feedback Submission
• Best Written Blog Post: write an optional blog post on GitHub