Senior Computer Vision Engineer

Augmodo•United States, CA

7d•$170,000 - $200,000

About The Position

Augmodo is at the forefront of spatial computing, where mapping connects the digital and physical worlds. We build real-time spatial systems that give brick-and-mortar retail unprecedented physical intelligence. We are seeking a tenured Computer Vision Engineer to serve as a technical thought leader alongside our engineering leadership. In this role, you will bridge advanced R&D with production engineering to solve some of the toughest, unsolved computer vision problems in complex retail environments. This role requires a combination of rigorous academic grounding (deep research experience/publishing) and battle-tested industry execution. You will spend roughly 50% of your time on hypothesis-driven R&D and evidence-based analysis and 50% on concrete implementation, model optimization, and edge deployment.

Requirements

8+ years of substantial industry experience delivering production-grade computer vision systems, paired with strong academic/research experience (Master's or Ph.D. level work in CS, Robotics, Electrical/Computer Engineering, or a related field).
Proven history of taking complex R&D/fundamental research and translating it into evidence-backed, production-grade pipelines.
Deep expertise in PyTorch, custom deep learning/graphics architectures, and Vision-Language Models (VLMs).
Hands-on experience with TensorRT, CUDA optimization, and low-level GPU acceleration.
Deep familiarity with Docker for reproducible ML environments.
Production-grade Python mastery; strong C++ capability for high-performance components.
Experience with dense image registration/warping algorithms.
Proven structured analytical approach to innovating in computer vision

Responsibilities

Design, implement, and optimize state-of-the-art deep learning, computer vision, and visual-language models (VLMs) for real-world retail edge deployments.
Bridge exploratory research and production code, executing data-backed analysis to validate algorithm performance against real-world engineering risks.
Optimize neural network inference for high-throughput, low-latency execution using TensorRT and custom CUDA primitives.
Containerize and deploy robust ML pipelines using Docker across cloud and edge/embedded environments.
Embrace modern AI-assisted workflows (Copilot, code-assisted tooling) to maximize engineering throughput and team productivity.