Full-Time

Research Engineer

3D Reconstruction and Rendering

Posted on 10/31/2025

World Labs

World Labs

51-200 employees

Generates interactive 3D scenes from images

No salary listed

San Francisco, CA, USA

In Person

Category
AI & Machine Learning (2)
,
Required Skills
Python
Tensorflow
CUDA
Pytorch
Machine Learning
Requirements
  • 3+ years of experience in 3D reconstruction, graphics, or applied machine learning roles, ideally within startup or fast-paced research environments
  • Extensive experience with machine learning frameworks such as PyTorch or TensorFlow, especially in the context of generative models and 3D representations
  • Deep expertise in at least one area of computer graphics, 3D reconstruction, neural rendering, generative AI for vision, or related fields
  • Proven ability to deliver high-quality 3D projects that involve complex object and scene reconstruction and/or rendering
  • Strong academic track record or demonstrated expertise in 3D understanding, machine learning and/or rendering
  • Strong coding proficiency in Python and experience with GPU-accelerated computing
  • Ability to engage effectively with researchers and cross-functional teams, clearly translating complex technical ideas into actionable tasks and outcomes
  • Comfortable operating within a dynamic startup environment with high levels of ambiguity, ownership, and innovation
Responsibilities
  • Develop and implement state-of-the-art 3D representations and algorithms to accurately capture and render detailed 3D structures
  • They need to work on small to large-scale objects and complex scenes, applying computer vision, graphics, and neural techniques
  • Collaborate closely with research and product teams to understand and translate product requirements into effective technical roadmaps
  • Contribute hands-on to all stages of development including data curation, experimentation, evaluation, and deployment
  • Continuously explore and integrate cutting-edge research in 3D reconstruction and generative AI
  • Act as a key technical resource within the team, mentoring colleagues, and driving best practices in 3D reconstruction and ML engineering
Desired Qualifications
  • Contributions to open-source projects in the fields of computer vision, graphics, or ML
  • Familiarity with large-scale training infrastructure (e.g., multi-node GPU clusters, distributed training environments)
  • Experience integrating machine learning models into production environments
  • Experience writing custom graphics or rendering related CUDA operators for use in ML frameworks such as PyTorch

World Labs is building tools for spatial intelligence through Large World Models (LWMs) that can perceive, generate, and interact with 3D environments. Its main product is an AI system that can turn a single image into an entire interactive 3D scene, useful for gaming, simulation, and digital content creation. The technology works by taking a 2D image and generating full 3D representations and interactions, forming a platform or service for developers to create 3D content and experiences. Unlike many competitors, World Labs focuses specifically on zero-shot 3D scene generation from a single image and providing end-to-end 3D content creation capabilities as a scalable platform. Its goal is to enable rapid creation and manipulation of rich 3D worlds for developers and companies in gaming, simulation, and digital media.

Company Size

51-200

Company Stage

Late Stage VC

Total Funding

$1.2B

Headquarters

Stanford, California

Founded

2024

Simplify Jobs

Simplify's Take

What believers are saying

  • Autodesk's $200M investment enables AEC industry collaboration since February 2026.
  • $230M funding from Andreessen Horowitz and Nvidia reached $1B valuation in 2025.
  • Google Cloud partnership provides scalable compute and enterprise distribution.

What critics are saying

  • NVIDIA Omniverse erodes gaming differentiation within 6-12 months.
  • Google DeepMind's Genie 2 surpasses Marble in agentic reasoning by mid-2027.
  • Autodesk integrates and commoditizes World Labs' IP in 12-24 months.

What makes World Labs unique

  • World Labs' Marble generates exportable 3D Gaussian splat assets from single images.
  • Large World Models predict physics, lighting, and spatial interactions beyond 2D language AI.
  • Fei-Fei Li's spatial intelligence focuses on persistent, editable 3D environments.

Help us improve and share your feedback! Did you find this helpful?

Benefits

Hybrid Work Options

Growth & Insights and Company News

Headcount

6 month growth

-20%

1 year growth

-19%

2 year growth

-8%
NextTech Today
Mar 16th, 2026
Autodesk invests $200M in World Labs to develop spatial AI for the physical world

Autodesk has made a $200 million strategic investment in World Labs, the spatial AI startup co-founded by Stanford computer scientist Dr Fei-Fei Li. The investment gives Autodesk an advisory role and close collaboration on research focused on physical-world AI systems that understand space, structure, materials, physics and time. World Labs develops multimodal world models designed to understand and generate realistic three-dimensional environments. Autodesk said this aligns with demands from its core industries—architecture, engineering, construction and manufacturing—where designing requires AI that can reason in three dimensions and support iterative workflows. CEO Andrew Anagnost said the investment represents a departure from AI spending focused on ever-larger models. The partnership aims to develop AI that understands geometry and physics rather than language alone.

World Labs
Feb 19th, 2026
World Labs Announces New Funding

An update on our vision for spatial intelligence in 2026.

PR Newswire
Nov 20th, 2025
Cisco Invests in Spatial Intelligence Pioneer World Labs

World Labs is redefining what's possible by giving AI the ability to generate, reason within, and interact with 3D worlds News Summary Cisco Investments is...

Bloomberg L.P.
Sep 13th, 2024
AI Pioneer Fei-Fei Li Raises $230 Million for New Startup World Labs

Andreessen Horowitz, Ashton Kutcher and Nvidia are among the investors.

Swipeline
Aug 15th, 2024
World Labs secures $100M funding

World Labs, founded by Stanford Professor Fei-Fei Li, secured $100 million in funding at a $1 billion valuation in a round led by NEA. The company completed two funding rounds within two months, with the first in April involving Andreessen Horowitz and Radical Ventures. World Labs, which became a unicorn four months post-establishment, develops AI models for gaming and robotics that can create detailed digital replicas of real-world objects and environments without extensive data.

INACTIVE