Software Engineer - CV Tooling

Augmodo•United States, CA

9d•$150,000 - $190,000

About The Position

Augmodo is building the "operating system for the physical shelf" using spatial computing and wearable AI. We provide real-time, store-level insights that were previously impossible to capture. We are moving fast, and we need an applied engineer who combines real Computer Vision expertise with high-velocity front-end tooling to make complex spatial data instantly intuitive and highly actionable. The Challenge We are looking for a Software Engineer to own the end-to-end user interfaces, active learning agents, and visual tooling powering our Spatial Platform. This is not just a front-end tools role—it requires a real Computer Vision background. You will leverage your experience building and deploying shippable CV models to build pre-labeling agents, custom models for active learning, functional editors, and model-training environments. Because you understand what it takes to train production-grade ML models, you know that high-quality data is everything. You will apply that knowledge directly to building optimized, intelligent annotation and curation tools for our internal researchers, algorithms, and external clients.

Requirements

Computer Vision Background: Hands-on experience building, training, and deploying real-world CV models to production. You understand end-to-end model lifecycles and know what "high-quality data" actually looks like.
Full-Stack Tooling: Strong expertise in React and TypeScript for interactive front-end tools, combined with deep capability in Python (PyTorch/TensorFlow, OpenCV) for the backend model integration.
Tooling & Human-in-the-Loop (HITL) Focus: Demonstrated track record building functional editors, canvas-based systems, active learning loops, or annotation tools powered by embedded ML/CV.
Bias for Action: High-velocity execution style. You are action-oriented, love leveraging modern AI tools (Cursor, Copilot, etc.), and prioritize getting working tools into researchers' hands fast.

Nice To Haves

Autonomous Systems / Spatial Data: Prior experience in self-driving data curation pipelines, spatial mapping, LIDAR, or 3D bounding workflows.
Retail/Logistics Domain: Experience handling SKUs, physical inventory maps, visual shelf data, or complex product catalogs.
3D/Spatial Web Tech: Familiarity with Three.js, WebGL, or GIS libraries for rendering spatial metadata.

Responsibilities

Applied CV for Tooling & Automation: Train and integrate lightweight, custom CV models, active learning workflows, and pre-labeling agents specifically designed to accelerate human annotation and data quality control.
Data Curation & Annotation Systems: Build robust internal QC, mapping, and annotation platforms (akin to autonomous vehicle curation pipelines) backed by an intimate understanding of CV data quality and edge cases.
Spatial & Planogram Interfaces: Design high-performance front-end interfaces that map raw product images and computer vision metadata directly to physical coordinates on digital retail planograms.
High-Velocity Iteration: Build functional editors, POCs, and pre-labeling pipelines rapidly to support active model development. We value high agency—shipping high-quality features in days, not months.
Consumer-Facing Simplicity: Translate complex, multi-layered spatial and ML outputs into simple, intuitive visual workflows for non-technical users.