ML Research Intern: 3D Scene Reconstruction for Autonomy

Rivian; Palo Alto (CA); On-site; USD 45,000 - 70,000; Full time

Benefits offered by this job

Paid vacation

Paid sick leave

Medical insurance

Job summary

Rivian invites a Machine Learning Research Intern to join the Autonomy team in Palo Alto. You will advance neural scene reconstruction and world modeling that turn real-world driving data into high-fidelity, re-renderable 3D environments.

This role emphasizes research, collaboration, and impactful publications. The internship runs Spring 2027, requires full-time onsite work, and aims to push the AI stack powering Rivian’s autonomous driving systems.

Qualifications

• Must be currently pursuing a PhD degree at an accredited university.
• Active student enrolled in CS/EE or a closely related quantitative field.
• Strong record of first-author publications in top ML/vision venues.

Responsibilities

• Research, prototype, and train 3D ML models for scene reconstruction and neural rendering.
• Design evaluation methods for reconstruction and rendering quality.
• Scale training and inference pipelines on large multi-GPU datasets.
• Collaborate with perception, simulation, and infrastructure teams to integrate prototypes.
• Present results and contribute to internal technical documentation.
• Deliver a final presentation summarizing research contributions.

Skills

Python

PyTorch

3D reconstruction

Distributed training

Multi-GPU

Open-source

Public portfolio

Education

PhD candidate

CS/EE or related field

Tools

PyTorch

Python

CUDA

Git

Job description

Rivian invites a Machine Learning Research Intern to join the Autonomy team in Palo Alto. You will advance neural scene reconstruction and world modeling that turn real-world driving data into high-fidelity, re-renderable 3D environments.

This role emphasizes research, collaboration, and impactful publications. The internship runs Spring 2027, requires full-time onsite work, and aims to push the AI stack powering Rivian’s autonomous driving systems.