Research Scientist – Vision-Language Models

About Arcade

Arcade is building the world's first AI physical product creation platform, enabling anyone to design, purchase, and sell custom, manufacturable products using natural language and generative AI. The platform democratizes physical product creation, making it as accessible as posting online. Arcade has raised $42M from leading investors including Reid Hoffman, Forerunner Ventures, Canaan Partners, and Sound Ventures. The company is headquartered in San Francisco's Presidio and led by serial entrepreneur Mariam Naficy (Minted, Eve). The Chief AI Officer is Varun Jampani, co-author of Dreambooth and creator of Stable Diffusion 3.5. The leadership team includes talent from Google, Apple, Stability AI, NVIDIA, Databricks, Stanford, MIT, and Berkeley.

The Role

We seek a high-caliber, deeply innovative Vision-Language Model (VLM) expert to lead efforts in teaching foundation models to evaluate consumer products like expert appraisers. This is an invention role, not standard implementation. You will design novel architectures, author proprietary training paradigms, and engineer systems that estimate abstract product qualities including rich context-aware captions, precise market price predictions, and subjective aesthetic quality scores. If you pioneer multimodal AI solutions and want your inventions powering a groundbreaking production platform, we want to hear from you.

Key Responsibilities

• Innovation & Invention: Pioneer novel neural architectures, loss functions, and multimodal integration techniques to solve unprecedented challenges in visual product perception.
• VLM Fine-Tuning & Adaptation: Lead deep fine-tuning, adaptation, and structural optimization of vision-language models for production-scale deployment.
• Multimodal Problem-Solving: Go beyond object detection to engineer systems predicting market prices, generating rich product captions, and evaluating aesthetic quality.
• Research Leadership: Conduct cutting-edge research, publish findings, and establish Arcade as a thought leader in applied multimodal AI.

About You

You are an expert in vision-language models with proven experience fine-tuning and adapting foundation models. You have strong fundamentals in deep learning, computer vision, and natural language processing. You thrive on unsolved multimodal problems and can invent technologies when existing solutions fall short. You value impact and want your research deployed at scale in production.

Compensation Disclosure

Editorial Market Estimate (Not Employer-Provided): The employer did not publish compensation. This estimate of $80K–$130K USD/year is based on jobsremotelist.online market research for similar roles and may vary significantly by experience, location, and qualifications.