🛠 Tech Stack

💼 About This Role

You'll work at the frontier of vision-language models, LLMs, and agentic systems applied to teaching AI to understand home designs. You'll own a distinct research thread from problem formulation through publication.

🎯 What You'll Do

• Build data pipelines to extract and curate structured datasets from design data.
• Design and implement models for vision-language understanding and structured output generation.
• Implement training, fine-tuning, and post-training pipelines including preference optimization.
• Develop evaluation suites and error taxonomies for targeted tasks.

📋 Requirements

• Active enrollment in a PhD program in CS, ML, or related field at a U.S. institution.
• Strong Python skills and hands-on experience with PyTorch training workflows.
• Solid understanding of at least two of: computer vision, transformers, LLMs, representation learning.
• Demonstrated research ability through publications, preprints, or open-source research code.

✨ Nice to Have

• Experience with vision-language models or multimodal foundation models.
• Experience building or fine-tuning agentic LLM systems that reason over structured state.
• Familiarity with post-training techniques: RLHF, DPO, or preference optimization.

🎁 Benefits & Perks

• 🏖️ Unlimited PTO
• 🏥 Comprehensive medical, dental, and vision coverage
• 📈 Significant equity in a well-funded startup
• 👶 Meaningful maternity/paternity leave
• 🏠 Home office stipend