AI Researcher

Block is one company built from many blocks, all united by the same purpose of economic empowerment. The blocks that form our foundational teams — People, Finance, Counsel, Hardware, Information Security, Platform Infrastructure Engineering, and more — provide support and guidance at the corporate level. They work across business groups and around the globe, spanning time zones and disciplines to develop inclusive People policies, forecast finances, give legal counsel, safeguard systems, nurture new initiatives, and more. Every challenge creates possibilities, and we need different perspectives to see them all. Bring yours to Block.

Job Description

Dessa specializes in audacious deep learning moonshots at Block. We choose projects that are impactful and transformative, not projects that are incremental improvements over the status quo. We have a strong focus on open source and are committed to sharing our work. Our mission is to push the boundaries of what’s possible with AI in order to benefit humanity.

We are seeking a Deep Learning Research Scientist who is passionate about expanding the limits of deep learning capabilities. As a Research Scientist at Dessa, you will create novel deep learning systems with the capability to perform tasks that were once considered unattainable. You will be fully remote and report into our Dessa AI Research Lead.

You Will:

• Help design and develop novel generative AI systems within Block
• Develop new ideas to improve existing deep learning systems
• Advance the research agenda of the team by building deep learning systems
• Create things that will be open to the world, allowing you to share your work with a global community through open source

You Have:

• Minimum Master’s degree in Computer Science or other quantitative fields; or equivalent experience working in an AI research lab
• Publications or a demo page to share that highlight a proven track record of exploratory research/work in the field of deep learning
• 5+ years of hands-on deep learning experience
• Strong background in deep learning theory and techniques, with hands-on experience in developing and training neural networks using frameworks such as PyTorch, JAX or TensorFlow
• Experience writing performant code and implementing high-performance deep learning algorithms
• Proficiency with Python
• Experience utilizing theoretical and empirical research to solve problems

We are excited about individuals who have made notable contributions in the following areas:

• LLMs
• Audio/Image/Video generation
• Multimodal models
• Agents/RAGs
• CUDA