BSD is looking for researchers to assist with the development and deployment of a GET3D model pipeline using the Nvidia stack. This project will require students to work alongside our machine learning team to implement an MLOps pipeline to take images and run them through training infrastructure to generate AI-created 3D models. The end goal will be a simple web interface to upload photos and a way to retrieve the models from the system.

This project is purely an experiment for testing infrastructure and will not be a commercial product at the end of the project.

Students will:

1. Create the automation pipelines using prefect 2.0 to automate the training process
2. Install cuda libraries and implement the open source GET3D tooling
3. Create a simple web interface to view pipeline activities and upload photos to generate the 3D models
4. Monitor the inference process and learn about how the tooling works.

Students will work directly with our machine learning team to get direct support through our chat channels on slack. We will meet regularly with the students to provide support. Our IT team will also be on this project to support them in accessing the technology on-site to remotely connect to the servers for training and running this technology.

BSD is a member of the Nvidia Inception accelerator, so students will also have access to free DLI training during this process, many of which include a certificate for completing the training.