Perception Lead (AI)

Location: San Francisco, California, United States; on-site

Base salary: $220,000–$260,000 per year, plus equity.

Company

hirify.global is incubating Sauron, a home security company developing an AI-powered perception system that combines sensors, computer vision, and nonlethal deterrence to protect homes and property.

What you will do

• Own the perception architecture and set technical direction across perception, cloud, product, and hardware.
• Define sensing requirements and accuracy, latency, cost, and readiness targets for successive hardware generations.
• Evolve media-pipeline architecture to support changing cameras and configurations without disrupting live video or losing object identity.
• Build and improve computer vision systems for multi-object tracking, re-identification, sensor fusion, occlusion, poor lighting, and deliberate evasion.
• Develop evaluation datasets, regression harnesses, deployment-data feedback loops, observability, failure detection, and graceful degradation.
• Lead technical work and people through architecture decisions, code reviews, collaboration, and ownership of perception interfaces.

Requirements

• 5+ years building production systems, including experience owning architecture at staff scope.
• Professional experience with perception or machine learning for hardware products in safety-critical fields such as aerospace, robotics, medical devices, autonomous vehicles, or physical security.
• Deep modern C++ experience, including C++20, lifetime management, threading, shutdown semantics, Abseil, gRPC, and CMake.
• Production experience with GStreamer or media pipelines, including pads, probes, caps negotiation, bus messages, and pipeline failure diagnosis.
• Shipped applied computer vision systems and strong knowledge of linear algebra, optimization, statistics, algorithms, and deep-learning lifecycles.
• Experience with Python, GPU inference, edge compute, TensorRT, batching, FP16, and inference-latency analysis.

Nice to have

• NVIDIA Jetson production experience, including JetPack, L4T, Yocto, or cross-building for aarch64.
• Multi-camera tracking, video surveillance, VMS, ONVIF/RTSP, CUDA, or GPU architecture experience.
• Familiarity with vision-language and other multimodal models for semantic scene understanding.
• Experience owning model evaluation datasets, metrics, and production-oriented model selection.

Culture & Benefits

• Collaborative environment with pair programming and teamwork.
• Ownership across the stack, with regular refactoring and test-driven development.
• Opportunity to shape a new perception system for physical security hardware.
• Equity included in the compensation package.
• Commitment to building a diverse and inclusive workforce.