Lead Agentic AI Engineer – Connectivity

Caterpillar Inc.•Irving, TX

1d•$128,470 - $192,710•Onsite

About The Position

The Lead Agentic AI Engineer defines how AI agents are engineered and deployed across Caterpillar’s connected-product ecosystem. As Connectivity’s hands-on technical lead for agentic AI, this role researches emerging technologies, rapidly prototypes approaches, and turns successful ideas into reusable engineering frameworks, reference architectures, and secure, enterprise-grade production solutions. The role works across agentic AI, software engineering, IoT, telematics, and connected machines to improve how engineers build, diagnose, support, and operate connected products. The engineer partners with Connectivity, Engineering, Cybersecurity, Enterprise AI, and business stakeholders throughout the solution lifecycle, from use-case discovery and architecture to integration, evaluation, deployment, observability, governance, and value realization. This is a hands-on engineering role involving code, prototypes, architectural decisions, and production engineering—not solely prompt engineering or research. Solutions must align with enterprise standards for security, compliance, performance, and responsible AI. Tools and platforms may include GitHub Copilot, models and platforms from Anthropic, OpenAI, xAI, other providers, and locally hosted or open models where appropriate. The position is based onsite at Caterpillar’s Irving, Texas office five days per week, requires up to 10% domestic travel, and offers domestic relocation assistance. Visa sponsorship is not available. The stated summary pay range is $128,470 to $192,710.

Requirements

Knowledge of application design and architecture principles and practices, including the ability to translate business requirements and logical models into technical application designs.
Ability to design and build secure, maintainable, production-grade AI agents and services in Python, including RAG, tool calling, context engineering, memory, MCP integration, multi-agent workflows, and human-in-the-loop controls.
Experience evaluating emerging LLMs, agent frameworks, protocols, and locally deployable models and developing prototypes with leading AI tools and platforms.
Experience taking software or AI solutions from prototype through production deployment and support.
Knowledge of systems thinking and the interdependencies among system elements, with ability to design solutions that synchronize resources to achieve business results.
Ability to apply AI to connected assets, telematics, embedded systems, device lifecycle management, remote services, and engineering workflows.
Ability to integrate enterprise systems, APIs, documentation, and connectivity data while considering identity, access, security, privacy, and data protection.
Ability to create automated tests, tracing, and observability for agent behavior, retrieval quality, latency, and reliability.
Ability to assess AI models and solutions for capability, performance, security, reliability, cost, and enterprise readiness.
Familiarity with cloud deployment, software testing, source control, and production monitoring.
Knowledge of business intelligence and the use of business processes and technologies to gather, store, analyze, and provide access to data for decision-making.
Ability to create evaluation datasets, acceptance criteria, and measurement approaches for AI solutions.
Ability to measure gains in engineering productivity, troubleshooting, service responsiveness, quality, time, and cost.
Ability to translate business problems into technical solutions and communicate results to technical and nontechnical stakeholders.

Nice To Haves

Deep hands-on experience designing agentic AI systems with LangGraph, LangChain, Semantic Kernel, CrewAI, or similar frameworks, including tool calling, RAG, memory, MCP, multi-agent orchestration, and human-in-the-loop patterns.
Experience evaluating, optimizing, and deploying open-weight or locally hosted LLMs, including model selection, inference optimization, security, observability, latency, and lifecycle management.
Experience architecting and operating AI workloads on Azure, AWS, or GCP using Docker, Kubernetes, infrastructure-as-code, CI/CD, and production monitoring.
Experience designing LLM evaluation and observability systems, including retrieval and response quality evaluation, automated testing, tracing, latency measurement, hallucination detection, and reliability monitoring.
Experience integrating AI agents with enterprise APIs, data sources, authentication and authorization systems, and production services.
Experience with IoT, telematics, embedded systems, connected equipment, or industrial digital platforms, including MQTT, HTTPS, cellular, satellite, and device-to-cloud communication.
Experience rapidly evaluating emerging models and AI development platforms such as GitHub Copilot, Anthropic Claude, OpenAI models, xAI models, or comparable technologies and determining suitability for enterprise production environments.

Responsibilities

Research and evaluate emerging LLMs, agent frameworks, protocols, and locally deployable models; build prototypes and assess capability, cost, performance, security, and enterprise readiness to recommend a supported AI engineering stack.
Develop reusable components for agent orchestration, context engineering, RAG, memory, tool use, MCP integration, multi-agent workflows, and human-in-the-loop controls.
Design and build secure, maintainable, production-grade AI agents and services in Python, taking solutions from prototype through deployment and operations.
Apply AI to connected assets, telematics, embedded systems, device lifecycle management, remote services, and engineering workflows.
Integrate approved enterprise systems, APIs, documentation, and connectivity data while enforcing identity, access, security, privacy, and data-protection requirements.
Create evaluation datasets, acceptance criteria, automated tests, tracing, and observability for agent behavior, retrieval quality, latency, and reliability.
Measure AI solution impacts on engineering productivity, troubleshooting, service responsiveness, quality, time, and cost to guide continued investment.
Serve as Connectivity’s technical point of contact for agentic AI and establish reference architectures and engineering practices that teams can adopt consistently.
Partner with enterprise architecture, cybersecurity, privacy, legal, and AI governance teams to establish secure, auditable, and responsible AI patterns.
Translate business problems into technical solutions and explain results to technical and nontechnical stakeholders.
Travel domestically up to 10% of the time.