Member of Technical Staff, Real-Time Vision Systems

Member of Technical Staff (MTS), Real-Time Vision Systems

San Francisco, CA | Remote (US)

About OpenVector OpenVector is making cognitively-inspired real-time computer vision systems. We believe the current paradigm of brute-force, scale-to-win visual processing is hitting a wall when it comes to latency, power consumption, and adaptability in the physical world. Instead, we are building systems inspired by how the human brain actually processes visual data—efficiently, dynamically, and with deep contextual awareness. Our team is a tight-knit group of ML researchers, neuroscientists, and systems engineers working to deploy fundamentally new visual architectures into autonomous systems and edge devices.

About the Role As a Member of Technical Staff on the Vision team, you will bridge the gap between cognitive science and high-performance engineering. We are moving beyond standard CNNs and ViTs to build models that process spatiotemporal dynamics in real time. You will be responsible for designing, training, and optimizing these architectures so they can operate reliably in noisy, real-world environments with strict latency constraints.

Your work will not sit in a research paper; it will translate directly into the perception engines powering our core products.

Key Responsibilities:

• Design, train, and evaluate novel vision architectures inspired by cognitive science and computational neuroscience (e.g., predictive coding, continuous-time dynamics, attention-driven spatial sampling).
• Optimize these models for real-time inference, balancing extreme performance requirements with hardware and power constraints.
• Build and scale robust data pipelines for streaming video, event-based data, and complex spatiotemporal datasets.
• Develop internal evaluation frameworks to test model performance against dynamic, real-world edge cases.
• Collaborate closely with our hardware and systems teams to integrate your models onto target edge devices.

Minimum Qualifications:

• Strong software engineering fundamentals and fluency in Python and C++.
• Deep experience training and deploying models using PyTorch, JAX, or similar frameworks.
• Proven track record of building computer vision models that have shipped to production.
• Solid understanding of real-time systems, memory management, and performance profiling.
• Excellent communication skills and the ability to work effectively in a highly collaborative, fast-paced startup environment.

Preferred Qualifications:

• Background in neuromorphic computing.
• Top publication track record (NeurIPS, CVPR, ECCV, ICCV etc.).
• Experience with non-traditional vision sensors (e.g., event cameras / DVS) or non-traditional architectures (e.g., Spiking Neural Networks).
• Experience with edge deployment and hardware acceleration (CUDA, TensorRT, ONNX, embedded GPUs).
• Strong background in video processing, 3D computer vision, or optical flow.

Explore our comprehensive directory of visa sponsorship jobs from employers worldwide who are ready to sponsor talented international professionals.

• Experience dealing with long-horizon visual tasks or continuous learning systems.

Compensation The annual base salary range for this role is listed below. This does not include equity compensation (options), which forms a significant part of the total compensation package for this role, alongside comprehensive health and wellness benefits.

Annual Salary: $135,000 - $250,000 USD

Logistics

• Minimum education: Bachelor’s degree in Computer Science, Electrical Engineering, Physics, Cognitive Science, or an equivalent combination of education and experience.
• Location policy: We operate on a hybrid model. We expect team members near our SF office 2-3 days a week for whiteboard sessions and hardware integration. Exceptional remote candidates within US time zones will be considered.
• Visa sponsorship: We are able to sponsor visas for this role (e.g., H1-B transfers, O-1) and retain outside immigration counsel to handle the process for our candidates.

We encourage you to apply even if you do not meet every single qualification. We are building systems that require a convergence of different disciplines, and we know that strong candidates come from non-traditional backgrounds. Research shows that people from underrepresented groups often exclude themselves prematurely if they don't check every box. We urge you not to do that. If you are passionate about the intersection of cognitive science and real-world engineering, we want to hear from you.

Security Note: OpenVector recruiters will only ever contact you from an @openvector.com email address. We will never ask you for banking information, equipment fees, or sensitive personal data prior to a formal offer and onboarding process.

How we're different We are highly pragmatic about our research. While we draw inspiration from biology and cognitive science, our ultimate metric is real-world performance.