Research Engineer
$180k - $340k • San Francisco • FullTime
Posted 10mo ago
Remote Work Policy
On-site
Employment Type
FullTime
Categories
AI Research Engineer
About the job
As a Research Engineer at Gamma, you will build models for visual communication, focusing on teaching them to reason about spatial composition, hierarchy, and visual language. This role is at the intersection of research rigor and product impact, offering the opportunity to develop evaluation frameworks and training data for a field with limited existing resources. You will fine-tune vision-language models to ensure Gamma's 100M+ users receive exceptional design quality with every generation. Success in this role requires a combination of deep expertise in VLMs and multimodal modeling, a research-oriented mindset, comfort with ambiguity, and a keen eye for visual and design quality.
Responsibilities
- Fine-tune vision-language models for layout generation, design critique, and content-to-visual translation.
- Design evaluation frameworks and benchmarks for visual communication quality, addressing layout, typography, color, and information density.
- Lead proprietary data collection for visual design tasks to train models on design principles.
- Conduct rigorous experiments to understand model behavior and implement targeted improvements.
- Diagnose and resolve systematic failure modes in production output.
- Build tools and workflows to accelerate team iteration and validation.
- Collaborate with product and engineering teams to deliver scalable quality improvements.
Requirements
- Hands-on experience with vision-language models or multimodal modeling (training, fine-tuning, or evaluation).
- Experience with post-training techniques such as supervised fine-tuning and reinforcement learning.
- Proven track record of building evaluations for subjective or hard-to-measure qualities.
- 2+ years of experience building AI systems with production deployment of user-facing models.
- Master's or PhD in Computer Science, Machine Learning, or a related field, or equivalent research experience.
- Strong publication record at top-tier conferences (e.g., NeurIPS, CVPR, ACL).
Benefits
- Benefits & equity