Research Scientist

Research Scientist

As a Research Scientist, you will originate and drive the novel machine learning research directions that shape Roboflow's research agenda. Because of Roboflow's huge userbase, we have a unique visibility into how people are using computer vision in the real world, including in what circumstances existing methods are unable to solve user problems.

Our Research Team's mission is to introduce novel methods that generalize as well as possible across our userbase. This includes methods that learn well on diverse data, run efficiently on different hardware targets, and solve unique categories of problems. You will identify the most promising open problems in computer vision, develop original theoretical approaches and algorithms to address them, and publish your work at top conferences. You'll partner closely with our engineering team, who will take your research and translate it into production systems running at scale.

Who You Are

You are a pragmatic, experienced machine learning researcher who wants to be an important part of an exceptional team that focuses on using Roboflow's computer vision tools to impact and improve every industry. You have high agency and a bias toward action.

• Masters / PhD in AI, machine learning, computer vision, robotics, or similar technical field of study, or equivalent industry experience

• History of publication at conferences / journals (CVPR, NeurIPS, ICLR, etc.)

• Deep understanding of ML theory and a track record of turning open problems into novel, well-formed research directions

• Values scientific rigour and reproducibility

• Passionate about computer vision and solving problems users value

• Pragmatic approach to utility in practice (not just in theory)

• Cares about open source and the broader research/developer ecosystem

• Writes and communicates clearly — able to explain your work to the scientific community, the developer community, and non-research stakeholders inside Roboflow

Preferred but not required:

• Familiarity with the practical constraints of optimizing models for edge hardware and training models on distributed infrastructure

• Experience collaborating with engineering teams to move research from prototype to production