This senior computer vision role suits an experienced machine learning professional looking to build models for traffic safety systems in San Carlos. The position involves owning the full model-development lifecycle for object detection and tracking in real-world environments.
Role Overview
Obvio is hiring a Senior Computer Vision Engineer based in San Carlos to build machine learning systems that power solar-powered, AI-assisted traffic enforcement cameras. This individual contributor position focuses on creating reliable detection and tracking models for real-world traffic conditions to reduce reckless driving and prevent traffic fatalities.
Key Responsibilities
Build and enhance object-detection and multi-object-tracking models for pedestrians, vehicles, and other road users in difficult real-world settings.
Manage the data strategy for model quality, covering dataset versioning, sampling, labeling, hard-negative mining, edge cases, class imbalance, and deployment feedback.
Conduct experiments and ablations to separate true performance improvements from overfitting, data leakage, noisy labels, or unsuccessful field deployments.
Construct reproducible training pipelines featuring experiment tracking, model registries, configuration management, metric dashboards, artifact lineage, and hyperparameter search.
Formulate evaluation strategies focused on product behavior, precision and recall balances, scenario slices, tracking quality, and calibration.
Collaborate with embedded engineers to adapt models for edge deployment while managing memory, latency, accuracy, and power limits.
Guide technical direction, host design reviews, mentor team members, and improve standards for production readiness and machine learning reproducibility.
Requirements
Seven or more years of professional experience in computer vision or machine learning, with a history of deploying and refining production models.
Extensive hands-on background in multi-object tracking and object detection, incorporating modern architectures, evaluation methods, and data augmentation.
Knowledge of domain shift, overfitting, dataset leakage, label noise, class imbalance, model calibration, and valid statistical validation.
Background in developing training platforms or infrastructure for distributed training, reproducible experiments, hyperparameter search, and model lineage.
Proficiency in Python and PyTorch alongside practical work handling large video and image datasets.
Experience verifying models using field data and managing the cycle from failure analysis to retraining and verified upgrades.
Demonstrated technical leadership in ambiguous, cross-functional projects with strong ownership and communication.
Nice to Have
Background in video-analytics domains, automotive, traffic, robotics, or surveillance.
Experience handling trajectory modeling, re-identification, occlusion, camera calibration, or multi-camera tracking.
Familiarity with synthetic data, active learning, weak supervision, automated labeling, or dataset quality tools.
Background optimizing and running vision models on edge accelerators like NVIDIA Jetson or Qualcomm Snapdragon.
Skills and Tools
Python
PyTorch
NVIDIA Jetson
Qualcomm Snapdragon
Object detection models
Multi-object tracking models
Distributed training infrastructure
About the Company
Obvio deploys solar-powered, AI-assisted cameras to automate traffic enforcement in vulnerable pedestrian areas, backed by Khosla Ventures and Bain Capital Ventures. The organization was founded by the team behind Motive's AI dashcam and secured a Series A of $54 million.
How to Apply
Interested candidates should apply directly on the official company careers page.
