Product Artificial Intelligence Engineer

📋 Description

• • Own the end-to-end lifecycle of AI/ML products that power the world’s toughest trucks—from 4×4 fire apparatus to hybrid military vehicles—ensuring every model you ship keeps soldiers, firefighters and entire communities safer.
• • Design, train and deploy predictive, computer-vision and NLP models that turn terabytes of telematics, CAD drawings and real-time sensor data into actionable intelligence: battery-life forecasts that prevent mission-critical failures, vision systems that guide autonomous material handlers, and generative dashboards that let engineers ask plain-language questions of their PLM vaults.
• • Partner with engine, hybrid, e-motor and suspension domain experts to embed AI into ECUs, cloud gateways and edge devices, balancing inference latency, thermal limits and functional-safety requirements while meeting MIL-STD and NFPA compliance.
• • Build automated data pipelines that ingest CAD files, CAN-bus logs, 3D scans and BOMs; use CV/OCR layout parsing to extract dimensions, GD&T tolerances and part numbers; then validate against ISO/ANSI standards and flag discrepancies before metal is cut.
• • Champion MLOps best practices: containerize models with Docker, orchestrate CI/CD through GitHub Actions or Azure DevOps, monitor model drift in production, and roll back gracefully when real-world conditions diverge from training data.
• • Translate complex model outputs into intuitive UX/UI elements—heat maps, confidence scores, natural-language explanations—so non-technical operators can trust and act on AI recommendations in seconds.
• • Collaborate with product managers to translate voice-of-customer insights into AI road-maps, estimate ROI, and prioritize features that reduce warranty claims, cut fuel burn or accelerate design cycles.
• • Ensure every algorithm meets Oshkosh’s ethical-AI and cybersecurity standards: bias audits, explainability reports, threat-model reviews and documentation that satisfy both internal governance and external regulators.
• • Generate production-grade Python, C++ and JavaScript templates that automate repetitive CAD, CAE and PLM workflows—complete with unit tests, linting and style guides—so engineering teams can iterate ten times faster.
• • Stay ahead of the curve: evaluate emerging frameworks (PyTorch 2.x, TensorFlow Lite, LangChain, LlamaIndex, vector stores, reinforcement-learning libraries) and run quarterly tech spikes that demonstrate tangible value to vehicle programs.
• • Mentor junior data scientists and interns, conduct code reviews, and foster a culture where experimentation is safe, knowledge is shared and every sprint ends with a demo that makes the whole room say “wow.”
• • Travel occasionally to proving grounds, customer sites or manufacturing plants to collect edge-case data, validate model performance under dust, vibration and extreme temperatures, and gather direct feedback from drivers and technicians.
• • Document everything—architecture decision records, API specs, training playbooks—so future teams can reproduce, extend and harden your work long after you’ve moved on to the next breakthrough.
• • Deliver results on aggressive yet realistic timelines, navigating shifting priorities with the agility of a startup while leveraging the scale, safety culture and global reach of a 100-year-old industrial leader.