Principal Software Engineer - Agentic AI & Simulation Automation

Caterpillar•Irving, TX

1d•Onsite

About The Position

Caterpillar is hiring a Principal Software Engineer to lead production-grade agentic AI and automation capabilities that help engineers create environments, configure experiments, execute simulations, analyze results, and make informed engineering decisions. The role supports construction autonomy and Physical AI through trusted digital twins, high-fidelity simulation, scalable GPU computing, intelligent automation, and rigorous validation. This is a hands-on technical leadership position with responsibility for architecture, critical implementation, engineering standards, cross-team alignment, and mentoring. The engineer will own a major technical capability across multiple teams and programs, turn ambiguous business and engineering problems into scalable architectures and production software, and balance research and prototyping with secure, maintainable delivery. The position partners with simulation, robotics, autonomy, controls, AI, data, cloud, product, and customer-facing organizations. The role is based in Irving, Texas (Dallas). Domestic relocation assistance and visa sponsorship are available. The listed salary range is $147,760–$240,110. Benefits are subject to plan eligibility, terms, and guidelines.

Requirements

Advanced software engineering experience using Python and one or more of C++, C#, Java, Go, or Node.js.
Experience with APIs, distributed systems, containers, automated testing, CI/CD, source control, and observability.
Experience with cloud platforms and cloud-native architectures, including Docker and Kubernetes.
Demonstrated ability to lead architecture and technical delivery across multiple teams.
Strong communication, mentoring, problem-solving, and cross-functional collaboration skills.
Experience using AI-assisted development tools such as GitHub Copilot, Claude, or Cursor.
Experience designing and delivering production agentic AI, generative AI, or intelligent automation systems.
Experience with NVIDIA NIM, NeMo, Nemotron, or comparable foundation-model and inference ecosystems.
Experience with NVIDIA Cosmos or comparable world foundation models for Physical AI, scenario generation, and simulation augmentation.
Experience connecting AI agents to NVIDIA Omniverse, Isaac Sim, OpenUSD, or other engineering simulation platforms.
Strong Python, API, distributed systems, event-driven workflow, and backend services experience.
Experience with RAG, embeddings, vector search, knowledge graphs, semantic data models, structured outputs, and multimodal models.
Experience with AI evaluation datasets, test harnesses, security controls, observability, and human-in-the-loop governance.
Experience deploying GPU inference using Docker, Kubernetes, NVIDIA GPUs, and CoreWeave or comparable GPU cloud infrastructure.
The candidate must be based in Irving, TX (Dallas).

Nice To Haves

Seven or more years of software development experience and three or more years developing ML, generative AI, or intelligent automation systems.
Experience delivering an AI-enabled capability from prototype through production.
Experience integrating agents with simulation, robotics, scientific computing, or enterprise engineering tools.
Working knowledge of Newton, physics simulation, and robotics workflows sufficient to orchestrate and evaluate tool use.
Familiarity with the preferred technology stack: NVIDIA Cosmos, NIM, NeMo, Nemotron, Omniverse, Isaac Sim, OpenUSD, agentic AI, RAG, Python, CUDA, Docker, Kubernetes, and CoreWeave.

Responsibilities

Define architecture for agentic simulation and multi-agent engineering workflows.
Build agents that interact with simulation engines, digital twins, APIs, datasets, and engineering tools.
Lead automated map-to-simulation, site-to-simulation, scenario-generation, and experiment-planning workflows.
Apply foundation models and multimodal AI to simulation configuration, monitoring, troubleshooting, analysis, and reporting.
Establish tool calling, workflow state, memory, human approvals, guardrails, observability, and auditability.
Create evaluation frameworks for accuracy, reliability, reproducibility, latency, cost, safety, and task completion.
Integrate AI services with NVIDIA Omniverse, Isaac Sim, OpenUSD, Cosmos, databases, knowledge systems, and cloud services.
Deploy and scale GPU inference and agent workloads on CoreWeave and Kubernetes.
Own a major technical capability across multiple teams and programs; translate ambiguous problems into scalable architectures and production software.
Lead architecture reviews, technical investigations, design decisions, and engineering standards.
Mentor senior and developing engineers and support technical hiring and onboarding.
Partner across simulation, robotics, autonomy, controls, AI, data, cloud, product, and customer-facing organizations.
Establish a reference architecture and technical roadmap for the capability area.
Deliver reusable platform components and measurable engineering outcomes within the first 12 months.
Improve quality, automation, scalability, and cross-team adoption while reducing duplicated solutions.
Develop technical depth across the broader engineering organization through mentoring and standards.