Senior Perception Engineer

About Sauron

Sauron is the home security company of the future. Leveraging cutting-edge AI, sensor technology, and nonlethal deterrence, Sauron brings next-generation technology to homeowners to protect their families and property. Incubated by Kevin Hartz and Atomic, Sauron has raised an $18M seed round from leading venture capital firms and angel investors, including 8VC and Flock Safety CEO Garret Langley.

The Role - Senior Perception Engineer

Perception is the foundation of the Sauron product. Every decision the system makes depends on whether we have correctly understood what took place outside the home.

Our hardware operates around the home and must complete its mission reliably in all environmental conditions: at night, in adverse weather, and in the presence of occlusion and deliberate evasion. In this role, you will set the technical direction for sensing, inference, and the perception architecture, balancing accuracy, latency, and cost tradeoffs. You will write a substantial amount of code focused on the most difficult problems and lead the technical direction across perception.

We Value

• Collaboration, pair programming, and teamwork.
• Taking ownership across the stack.
• Test-driven development, and refactoring regularly to keep codebases healthy.

Responsibilities

• Evolve the pipeline architecture so cameras and configurations can change dynamically without disrupting live video or losing object identity.
• Set the boundary between systems and models across compiled services and inference graphs.
• Own perception quality end to end, defining quantitative standards via evaluation data, regression harnesses, and hardware-appropriate model selection.
• Extract maximum value from sensors by fusing observations robust to occlusion, low lighting, and deliberate evasion.
• Advance services toward trustworthy, unattended operation with failure detection, graceful degradation, and observability.
• Close the feedback loop using deployment data to drive dataset and evaluation infrastructure improvements.
• Lead perception direction, review complex changes, and define system interfaces.

Qualifications

Required

• Around 5+ years of experience building production systems, with experience at staff scope owning system architecture.
• Significant experience with perception or machine learning for hardware products in a safety-critical field (aerospace, robotics, medical devices, autonomous vehicles, or physical security).
• Deep proficiency in modern C++ (C++20, Abseil, gRPC, CMake) with experience managing lifetime, threading, and shutdown semantics in daemons.
• Production GStreamer or media-pipeline experience (pads, probes, caps negotiation, bus messages).
• Applied computer vision experience shipping multi-object tracking, re-identification, or multi-camera association.
• Solid grounding in linear algebra, optimization, statistics, and algorithms.
• End-to-end deep learning lifecycle experience with PyTorch (or equivalent), custom ops/layers, edge inference optimization, reproducibility, and evaluation.
• GPU inference experience with TensorRT engines, batching, fp16, and latency analysis.
• Edge debugging skills and a strong focus on device-level observability and failure classification.
• Fluency in Python.
• Generalist mindset spanning cloud training infrastructure down to embedded systems.
• Excellent written and verbal communication skills.

Nice to Have

• Production experience with NVIDIA Jetson (JetPack, L4T, Yocto, aarch64 cross-building).
• Multi-camera tracking system experience.
• Video surveillance, VMS, or ONVIF/RTSP integrations.
• GPU architecture understanding and CUDA programming.
• Familiarity with Vision-Language Models (VLMs) and multi-modal models for semantic scene understanding.
• Experience owning model evaluation datasets and metrics.