Lead Data Scientist - Gen AI & Digital Twin

Caterpillar Inc.•Peoria, IL

About The Position

Your Work Shapes the World at Caterpillar Inc. When you join Caterpillar, you're joining a global team who cares not just about the work we do – but also about each other. We are the makers, problem solvers, and future world builders who are creating stronger, more sustainable communities. We don't just talk about progress and innovation here – we make it happen, with our customers, where we work and live. Together, we are building a better world, so we can all enjoy living in it. The Cat® Digital group is the digital and technology arm of Caterpillar Inc., responsible for bringing world class capabilities to our products and services. With over 1.5 million connected assets worldwide, we're focused on using data, advanced analytics, and AI capabilities to help our customers build a better world. To accomplish this, we’re deploying analytics that generate insights, recommend optimized decisions, and improve products by intelligently integrating massive quantities of telematics information, transactional records, images, unstructured documents, and other data sources. Job Summary: The Aftermarket Analytics (Condition Monitoring) team of Cat Digital is seeking a Lead Data Scientist to be a technical expert, working in a team environment, to support the development & integration of digital twins for condition monitoring & generative AI assisted predictive analytics for Caterpillar digital applications.

Requirements

Generative AI & LLMs: Proficiency in Fine-tuning and Prompt Engineering for Large Language Models, specifically using Retrieval-Augmented Generation (RAG)
Condition Monitoring Algorithms: Deep understanding of Anomaly Detection, Time-Series Analysis, and Predictive Maintenance models.
Telematics: Experience handling high-frequency IoT sensor data, CAN bus protocols (J1939), and integrating with unified data platforms
Experience with High performance computing
Business Statistics: Extensive experience with statistical tools, processes, and practices to describe business results in measurable scales; ability to use statistical tools and processes to assist in making business decisions.
Analytical Thinking: Extensive knowledge of techniques and tools that promote effective analysis; ability to determine the root cause of organizational problems and create alternative solutions that resolve these problems.
Programming Languages: Extensive knowledge of basic concepts and capabilities of applying Python programming to solve business challenges; ability to use tools, techniques and platforms in order to write and modify programming languages.
Requirements Analysis: Working knowledge of tools, methods, and techniques of requirement analysis; ability to elicit, analyze and record required business functionality and non-functionality requirements to ensure the success of a system or software development project.

Nice To Haves

Typically, a Bachelors, Masters, or PhD degree in Applied Statistics, Data Science, Business Analytics, Predictive Analytics, Business Intelligence & Analytics, Mathematics, Computer Science, Engineering (Aerospace, Electrical, Mechanical, Computer, Industrial, Agricultural, etc.), or equivalent technical degree
Extensive experience applying Python (NumPy, SciPy, pandas, etc.) programming to solve business challenges.
Extensive experience with advanced data analysis, machine learning such as clustering, Log regressions, neural nets and statistical methods such as statistical process control, etc. (typically 8+ years)
Experience in practical applications of onboard architecture / software (e.g. mini projects using Raspberry Pi or any other architecture is a bonus)
Working experience with heavy equipment engineering or data analysis.
Working knowledge with cloud technologies (AWS, Azure, Google Cloud, etc.)
Advanced experience with version control / repositories such as GitHub
Experience operating in an Agile environment
Must demonstrate strong initiative, interpersonal skills, and the ability to communicate effectively.

Responsibilities

Algorithm Development & Modeling
Anomaly Detection: Design and implement GPU-accelerated machine learning models (e.g., XGBoost, autoencoders, and GANs) to identify irregular patterns in high-frequency sensor data.
Digital Twin Engineering: Partner with engineering teams to develop onboard digital twins using NVIDIA architecture to simulate, predict, and optimize the performance of heavy machinery
Optimization: Profile and tune deep learning algorithms for maximum efficiency on NVIDIA GPU architectures, ensuring high throughput and low latency for real-time monitoring.
Testing onboard Architecture & Integration
Edge Deployment: Adapt and test algorithms for onboard architecture, leveraging tools like NVIDIA Jetson and real-time edge processing on Cat equipment.
Hardware-Software Co-Design: Collaborate with hardware / simulation engineers to ensure algorithm compatibility with next-generation processors and specialized onboard compute modules.
Simulation-Based Training: Use high-fidelity digital twins to simulate rare failure scenarios, ensuring the GenAI assistant provides accurate troubleshooting steps for edge-case mechanical issues.
GenAI Algorithm Automated Diagnostic Workflows: Develop Generative AI agents that synthesize telematics data to generate prioritized repairs for identified machine faults.
Unified Data Orchestration: Integrate multi-modal outputs from condition monitoring analytics & asset life history to create a machine-specific context for AI assistant.