About the Role
The role involves developing and refining high-performance algorithms to improve efficiency and scalability in 3D reconstruction and neural rendering pipelines.
Responsibilities
- Design and implement optimization techniques for machine learning models
- Improve computational efficiency of 3D reconstruction systems
- Collaborate with research and engineering teams to integrate performance improvements
- Analyze bottlenecks in existing pipelines and propose solutions
- Develop profiling tools to measure system performance
- Optimize memory usage and latency in real-time applications
- Translate research prototypes into production-grade code
- Evaluate trade-offs between accuracy and speed in model inference
- Work with large-scale datasets to train and validate models
- Contribute to the development of novel neural rendering methods
- Ensure software maintainability and scalability
- Benchmark performance across different hardware platforms
- Support deployment on cloud and edge devices
- Refine model quantization and compression strategies
- Collaborate on system architecture design
- Maintain up-to-date knowledge of advancements in optimization research
- Publish findings in top-tier conferences when applicable
- Write clean, testable, and well-documented code
- Participate in code reviews and technical discussions
- Assist in defining performance metrics and success criteria
- Troubleshoot and resolve performance regressions
- Integrate third-party libraries and tools where beneficial
- Optimize GPU and CPU utilization
- Work closely with product teams to align technical goals
- Ensure solutions meet real-world constraints
Nice to Have
- PhD in a relevant technical field
- Publications in machine learning, computer vision, or graphics venues
- Experience with real-time rendering engines
- Knowledge of differentiable rendering
- Familiarity with AR/VR applications
- Experience optimizing models for mobile platforms
- Background in mathematical optimization
- Contributions to open-source performance tools
- Prior work in neural radiance fields (NeRFs)
- Understanding of hardware constraints on embedded systems
Compensation
Competitive salary and equity package
Work Arrangement
Hybrid
Team
Small, cross-functional team focused on rapid iteration and technical innovation
About the Team
The team operates at the intersection of research and product development, focusing on making cutting-edge 3D technologies practical and performant.
What We Build
We develop systems that convert video inputs into high-fidelity 3D models using advanced machine learning and computer vision techniques.
Available for qualified candidates
