Machine Learning Engineer, Figma

• Expertise building infrastructure for AI/ML model development in industry
• Experience in software development in Python. Experience in languages like C/C++ is nice to have but not required
• Expertise working closely with AI researchers on applied AI projects to build high quality models in a fast paced environment
• Expertise improving datasets and evaluation metrics for applied ML projects
• Experience solving sophisticated problems and comparing alternative solutions, trade-offs, and diverse points of view to determine a path forward
• Experience communicating and working across functions to drive solutions

• We’re looking for AI engineers who have experience building data pipelines to collect high quality data, and evaluation systems to evaluate AI models
• You will be collaborating with AI researchers and product engineers closely to improve Figma’s AI features, build new capabilities and advance the state of the art in AI for creativity
• Build data pipelines to collect high quality training and evaluation data for developing AI models
• Build annotation pipelines to collect ground truth using humans and AI models in the loop
• Collaborate closely with AI product engineers to define requirements for AI features
• Collaborate closely with AI researchers to identify failure cases, improve data sets, evaluation metrics and hill climb quality of AI models
• Work closely with AI infra engineers to build the right infrastructure for AI model development and inference
• Explore the boundaries of what is possible with the current technology set and experiment with novel ideas

• Equity
• Health, Dental, & Vision
• Retirement with company contribution
• Parental leave & fertility support
• Mental health and wellness benefits
• Generous PTO
• Company recharge days
• Learning & development stipend
• Work from home stipend
• Cell phone reimbursement

Total funding: $749.1m

Figma is a web-based design platform that helps cross-functional teams of designers, project managers, and developers easily collaborate on product design projects. Its standout feature, Multiplayer mode, is like Google Docs for design, letting team members collaborate in real-time on project files that are hosted in the cloud.