Responsibilities
- Author and shepherd OT AI reference architectures and reusable design patterns through architecture and cybersecurity review.
- Define safety guardrails, human oversight, fail-safe behavior, degraded modes, rollback, explainability, and auditability for AI models and agents.
- Establish architecture patterns for edge AI, retrieval augmented generation, computer vision, predictive and prescriptive analytics, generative AI, agentic systems, and physical AI.
- Own OT AI model lifecycle architecture from intake and validation through deployment, monitoring, retraining, retirement, and rollback.
- Define requirements for AI training, validation, and inference data, including provenance, quality, labeling, contextualization, access controls, and secure movement.
- Evaluate AI vendors, platforms, and models for operational fit, safety, cybersecurity, supportability, interoperability, scalability, and lock-in exposure.
- Design testing and validation approaches for constrained, intermittently connected, or disconnected industrial environments.
- Produce AI risk assessments, architecture decision records, model documentation, data flow diagrams, and operational readiness evidence.
- Apply the Purdue model and ISA/IEC 62443 zone and conduit concepts to AI connectivity, edge inference, remote support, and model or data movement.
- Define observability requirements for AI services, including performance, confidence, drift, anomalous behavior, resource consumption, and operational impact.
- Partner with OT architecture, cybersecurity, operations, engineering, data science, Facility of the Future, and business unit teams to operationalize OT AI.
Requirements
- Bachelor’s degree in Engineering, Computer Science, Data Science, or a related field, or equivalent experience.
- At least 8 years of experience in operational technology, industrial automation, or industrial AI, including significant solution architecture responsibility.
- Experience architecting, deploying, or governing AI and machine learning solutions in operational or safety-adjacent environments.
- Experience with model lifecycle management, AI safety guardrails, human-in-the-loop controls, fallback behavior, and operational acceptance criteria.
- Experience designing edge AI solutions for constrained, intermittently connected, or disconnected industrial environments.
- Working knowledge of OT cybersecurity, the Purdue reference model, and ISA/IEC 62443 zone and conduit concepts as applied to AI systems.
- Working knowledge of Microsoft Azure AI and machine learning services and hybrid or on-premises deployment patterns.
- Ability to communicate AI design decisions, alternatives, risks, and consequences to technical, operational, cybersecurity, and business stakeholders.
- Demonstrated ability to evaluate AI vendors, platforms, and models and reach defensible technical verdicts.
- Preferred qualifications include a master’s degree, experience with generative AI, agentic AI, multimodal systems, computer vision, physical AI, robotics, drones, autonomous systems, MLOps, responsible AI, industrial datasets, semantic models, knowledge graphs, digital twins, edge AI accelerators, containerized inference, orchestration, observability, and industrial operations.
Benefits
- Relocation will not be considered.
- Expatriate assignments will not be considered.
Tech Stack
Categories
About Chevron
Our greatest resource is our people. Their ingenuity, creativity and collaboration have met the complex challenges of energy’s past. Together, we’ll take on the future. We support the LinkedIn Terms of Use (User Agreement), and we expect visitors to our page to do the same. We encourage open, lively conversation with a few simple rules: --We reserve the right to correct factual errors. --We will reply to comments when appropriate. --If we disagree with other opinions, we will do so respectfully. --You may not post anything that is spam or that is abusive, profane, or defamatory toward a person, entity, belief, or symbol. --We will delete any posts that contain personal information such as email addresses, phone numbers and physical addresses, and other third party intellectual property material, when that information does not belong to the author of the post. --You may not post job listings for non-Chevron positions. --While we support lively, open discussion, we reserve the right to delete comments.