🛠 Tech Stack

💼 About This Role

You'll design and execute research on multimodal model training with a focus on vision-language and speech-language models at Sword Health, an AI-native healthcare platform. You'll advance models for real-time patient understanding and treatment planning, contributing to both top-tier publications and production systems.

🎯 What You'll Do

• Design and execute research on multimodal model training including SFT and RLHF.
• Develop models for AI agents to perceive patients through video, language, and speech.
• Contribute to multimodal dataset curation, architecture design, and evaluation.
• Collaborate across teams to translate research breakthroughs into production systems.

📋 Requirements

• PhD in Computer Science, Machine Learning, NLP, Computer Vision, or related field.
• Hands-on experience fine-tuning LLMs or multimodal large models with SFT/RLHF.
• Experience training/fine-tuning models across multiple modalities (video, language, speech).
• Strong publication track record in peer-reviewed AI conferences or journals.

✨ Nice to Have

• First-author publications in top-tier AI conferences (NeurIPS, ICML, CVPR, etc.).
• Deep expertise in vision-language models or multimodal representation learning.
• Experience building LLM-based agents or deploying multimodal ML pipelines in production.

📨 Hiring Process

Estimated timeline: 2-4 weeks · AI estimate

1. 1Recruiter Screen· 30 min
2. 2Technical Interview· 60 min
3. 3Onsite / Final Round· half-day