AI Research Engineer – Multi‑Modal & Vision

About the role

This position is for an AI Research Engineer focused on multi‑modal systems that combine vision and language. Based in Switzerland, you will work within a highly skilled team to develop end‑to‑end AI solutions, from dataset creation to production deployment.

Key responsibilities

• Design, train, evaluate, and deploy vision‑language models.
• Implement post‑training techniques such as supervised fine‑tuning, knowledge distillation, and reinforcement learning from human feedback.
• Curate and maintain high‑quality multimodal datasets for specific domains.
• Optimize model efficiency for resource‑constrained environments using compression and adaptation methods.
• Build and manage distributed training workflows on GPU clusters, identifying performance bottlenecks.
• Develop benchmarking and evaluation frameworks to measure robustness and real‑world performance.
• Contribute to open‑source AI ecosystems and stay abreast of emerging research.

Required profile

• Bachelor’s degree in Computer Science, Machine Learning, AI or related field; Master’s or PhD preferred.
• Strong hands‑on experience with multimodal AI, especially vision‑language models.
• Proven track record of improving model performance on industry benchmarks or production use cases.

Required skills

• Supervised fine‑tuning, knowledge distillation, reinforcement learning from feedback.
• Parameter‑efficient fine‑tuning approaches.
• Distributed training frameworks and GPU infrastructure.
• Model optimization, compression, and adaptation techniques.
• Scalable machine‑learning pipeline development.