AI Automation Engineer – Data Annotation & Vision Systems

Caterpillar Brazil; Irving (TX); On-site; USD 98,000 - 158,000; Full time

Benefits offered by this job

Relocation assistance

Visa sponsorship

Job summary

Caterpillar Inc. is seeking an AI Automation Engineer to develop annotation automation solutions for robotics, autonomy, computer vision, and Vision-Language-Action systems.

You will join a scrum team focused on production-grade data pipelines and AI-driven labeling for real-world heavy equipment data. You will design labeled data workflows, train ML models, and integrate Caterpillar data with external annotation providers while ensuring reliability and scalability.

Qualifications

• Bachelor's degree in a relevant field and practical software engineering experience.
• Proven ability to design and deploy ML/AI data pipelines across teams.
• Experience with annotation quality and labeling guidelines.

Responsibilities

• Develop and deploy automated labeling and captioning solutions for perception and VLA datasets.
• Train, fine-tune, and evaluate AI models for auto-labeling and QA.
• Build data pipelines moving datasets through labeling, review, validation, and publication.
• Implement and extend platforms such as NVIDIA Cosmos Curator and Cosmos Reason.

Skills

Python

Docker

Kubernetes

APIs

Data pipelines

ML models

Data annotation

QA processes

Labelbox

Supervisely

RoboFlow

PyTorch

TensorFlow

ONNX

MLOps

LiDAR data

Education

Bachelor's degree in Computer Science, Software Engineering, AI/ML, Robotics

Tools

Labelbox

Supervisely

RoboFlow

NVIDIA Cosmos Curator

Cosmos Reason

Job description

Caterpillar Inc. is seeking an AI Automation Engineer to develop annotation automation solutions for robotics, autonomy, computer vision, and Vision-Language-Action systems.

You will join a scrum team focused on production-grade data pipelines and AI-driven labeling for real-world heavy equipment data. You will design labeled data workflows, train ML models, and integrate Caterpillar data with external annotation providers while ensuring reliability and scalability.