About the role#
LinkedIn is seeking PhD students for a Generative AI Engineering internship during the summer of 2027. You will join our AI/ML teams to advance the frontier of generative models and apply them to text, image synthesis, and multimodal learning. This role involves taking projects from initial ideation through to implementation, creating production-ready systems that improve the member experience.
What you'll need#
- Currently pursuing a PhD in computer science, statistics, mathematics, electrical engineering, or a related technical field.
- You must be returning to your PhD program after the internship, with an expected graduation date of December 2027 or later.
- Experience with Generative AI, such as LLMs, GANs, VAEs, or diffusion models.
- Proficiency in Python and deep learning frameworks like PyTorch, TensorFlow, or JAX.
- A background in neural networks and probabilistic methods.
- Knowledge of reinforcement learning, computer vision, and graphics problems is preferred.
- Familiarity with deploying generative models in production environments.
- Strong command of algorithms and data structures.
- Publications in AI/ML conferences like NeurIPS, ICML, CVPR, or ICCV are a plus.
What you'll do#
- Conduct research and development on generative models, including transformers, diffusion models, GANs, and autoregressive architectures.
- Apply techniques to text generation, creative content, conversational agents, and multimodal learning.
- Develop and implement large-scale, production-quality systems that integrate with the LinkedIn platform.
- Design evaluation frameworks, including metrics, datasets, and automated testing pipelines.
- Collaborate with product teams to build AI-driven user experiences.
- Contribute to internal frameworks for human-in-the-loop annotation and preference modeling.
Location & details#
- This position is based in Mountain View, California.
- The role is hybrid, requiring work from home and from the office on select days.
- This is a full-time, paid internship for Summer 2027.


