About Paperspace
Paperspace is a high-performance cloud computing and ML development platform for building, training and deploying machine learning models. Tens of thousands of individuals, startups and enterprises use Paperspace to iterate faster and collaborate on intelligent, real-time prediction engines.
Paperspace is backed by leading investors including Battery Ventures, Intel Capital, SineWave Ventures, Sorenson Ventures, Y Combinator and Initialized Capital.
Paperspace Advanced Technologies Group
Paperspace ATG is the internal R&D arm of Paperspace and is tasked with exploring advanced topics in machine learning/deep learning, data engineering, UI/UX for developing intelligent applications, and AI education/accessibility.
The Research Fellowship is a 10-15 week paid program that is designed to bring in Graduate and post-Graduate students from a wide variety of disciplines who want to apply their passion for research in a practical setting.
The ATG accepts fellows on a rolling basis. Research fellows are paired with a Paperspace engineer to advance and document their research project.
Research Areas
The ATG is currently engaged in the following active areas of research. Not all projects fall directly under any single area but generally, these areas encompass the types of projects that we would like to foster and support.
- Deep Tech - This research track is designed to push the limit of current machine learning algorithms, applications, or foundational knowledge. Past fellows have taken on GPU kernel development, adversarial auto-encoders, auto-ml parameter space exploration strategies, and other advanced topics in the deep learning field.
- Tooling / Interfaces - This research track is designed to push the boundaries of what is possible with ML interfaces (GUI, CLI, and others). Topics such as experiment tracking, visualization techniques, new distributed training architectures, and novel ways of modeling complex data and interactions fall under the scope of this research area. Past fellows have worked on making pre-trained deep learning models more accessible to a broader audience through new abstractions and interfaces. Model optimization for various inference architectures and device constraints is also an area of interest.
- Education / Accessibility - This research track is designed to expand the accessibility of existing ML and deep learning techniques through education and advocacy. As more and more advanced topics come from academia, there exists a growing divide between experts and novices. Furthermore, what are the best ways to open up deep learning to new audiences and users, not just domain experts? Questions of fairness, bias, explanation, openness are at the center of this conversation.
If you have any additional questions, you can reach out to atg@paperspace.com.
We are an equal opportunity employer that values and welcomes diversity. We are committed to building a team that represents a variety of backgrounds, perspectives, and skills. All qualified applicants will receive consideration for employment without regard to race, color, religion, gender, gender identity or expression, sexual orientation, national origin, genetics, disability, age, or veteran status.