Edge AI & Autonomy Runtime Architect

Koitecc Solutions; Irving, Northern (TX, KY); Hybrid; USD 148,000 - 240,000; Full time

Benefits offered by this job

Medical, dental, and vision benefits

401(k) with matching

Paid time off and holidays

Relocation assistance

Visa sponsorship available

Job summary

Caterpillar Inc. is seeking an experienced engineer to own the compute, OS, middleware, and runtime architecture for autonomy software on the NVIDIA Thor-based edge platform. You will enable perception, robotics and data logging, ensuring reliable edge operation.

Work will focus on GPU-accelerated inference, sensor data processing and runtime optimization, with emphasis on safety and reliability for real machines in autonomous construction sites.

Qualifications

• Strong experience with edge AI runtime platforms and embedded Linux systems.
• Familiarity with NVIDIA Jetson IGX/AGX Thor ecosystem is a strong plus.
• Experience with runtime health management, logging and updates on edge platforms.
• Ability to collaborate across perception, robotics, safety and cybersecurity teams.

Responsibilities

• Define edge compute and runtime architecture for autonomy services on the Thor platform.
• Leverage Jetson software stack (Linux, CUDA, TensorRT, Isaac) for inference and deployment.
• Define resource allocation (CPU/GPU/memory/storage/network) across services.
• Define runtime monitoring, health management, fault detection and diagnostics.

Skills

Edge AI runtime platforms

Embedded Linux systems

Robotics compute platforms

NVIDIA Jetson ecosystem

CUDA

TensorRT

Isaac

Resource management

Tools

Jetson Linux tooling

Middleware & orchestration

Job description

Caterpillar Inc. is seeking an experienced engineer to own the compute, OS, middleware, and runtime architecture for autonomy software on the NVIDIA Thor-based edge platform. You will enable perception, robotics and data logging, ensuring reliable edge operation.

Work will focus on GPU-accelerated inference, sensor data processing and runtime optimization, with emphasis on safety and reliability for real machines in autonomous construction sites.