Multimodal AI Research Intern — Foundation Models

Apple; Zürich; On-site; CHF 20,000 - 27,000; Full time

Job summary

Apple is seeking a PhD student or researcher to advance multimodal foundation models and agentic systems within the Multimodal Intelligence Team (MINT) in Zürich. You will explore post-training methods like SFT and RL to improve model capabilities and real-world integration into Apple products.

You’ll work with world-class engineers and scientists, publish research, and contribute to interdisciplinary efforts across vision, language, and multimodal understanding.

Qualifications

• Currently pursuing a PhD in ML, CV, NLP, data science or related areas.
• Experience with large language or vision-language models and their agentic applications.
• Proficient Python and at least one modern DL framework (PyTorch, JAX, or TensorFlow).

Responsibilities

• Advance post-training capabilities of multimodal foundation models for agentic applications.
• Design evaluation frameworks and training strategies to improve model perception, reasoning and autonomy.
• Work with large multimodal datasets and publish novel research at top-tier conferences.

Skills

PhD candidate

Python

PyTorch

JAX

TensorFlow

Education

PhD or equivalent experience

Tools

PyTorch

JAX

TensorFlow

Job description

Apple is seeking a PhD student or researcher to advance multimodal foundation models and agentic systems within the Multimodal Intelligence Team (MINT) in Zürich. You will explore post-training methods like SFT and RL to improve model capabilities and real-world integration into Apple products.

You’ll work with world-class engineers and scientists, publish research, and contribute to interdisciplinary efforts across vision, language, and multimodal understanding.