Our visual modeling team is currently seeking new members to focus on the advancement of vision-centric LLMs. We are building VLMs capable of perceiving intricate visual details and understanding user aesthetics to facilitate communication through visual assets using tools like multimodal search and text-to-image models
This role offers the opportunity to work with Pinterest’s unique visual-text datasets to develop large-scale generative models for production. You will join the core visual pod, a collaborative group of approximately six engineers and a product prototyping team, to create specialized evaluation benchmarks and contribute to the broader research community
Prototype new model architectures for Pinterest VLMs. We’re looking for hands-on experience working with finetuning open-source LLM models and improve their visual perception and tool using capabilities
Develop new evaluation benchmarks that tailors to vision-centric capabilities such as fashion style recommendations
Read research papers, participate in group discussions, and help brainstorm our overall visual generative strategy at the company
Help with collection of relevant visual training data for Pinterest Canvas, particularly to conduct RLHF, targeted fine-tuning, etc
Publish and publicize your work via conferences, paper submissions, blog posts, etc
Mentor more junior researchers or research interns within the Pinterest Labs organization
2+ years of industry computer vision experience
Research engineers and scientists who have experience working with generative computer vision models, preferably various forms of visual encoders and LLMs
M.S. or PhD in Machine Learning, Computer Science, or related areas
Publications at top ML conferences
Familiarity with LLM-powered productivity tools for documentation search, experiment analysis, SQL/data exploration, and engineering workflow acceleration
Experience using Cursor, Copilot, Codex, or similar AI coding assistants for development, debugging, testing, and refactoring