Principal AI Engineer

Job Description

Performs a combination of duties in accordance with departmental guidelines:

• Acts as one of principal engineers for CNA's AI-native engineering platform, designing the end-to-end system spanning agentic coding workflows, skills and agent marketplaces, AI-augmented CI/CD pipelines, automated quality gates, and rapid environment provisioning.
• Leads integration of AI tooling (Claude Code, Cursor, GitHub Copilot) into the software delivery lifecycle, ensuring these capabilities compose into a coherent, governed platform.
• Designs and builds the agentic infrastructure layer — including multi-agent orchestration patterns, sub-agent frameworks, skill authoring standards, and context engineering best practices — that enables engineering teams to operate at AI-native speed without sacrificing architectural integrity or security posture.
• Provides expert technical consultation to engineering leadership, portfolio teams, and architecture on how to adopt AI-native development practices, evaluate AI-generated code quality, and integrate agentic tooling into existing workflows.
• Advises on trade-offs between speed and quality, human-in-the-loop requirements, and appropriate levels of AI autonomy for different risk profiles (e.g., Sox-classified systems vs. rapid prototyping).
• Led the technical strategy for the centralized skills and agent marketplace, defining contribution standards, review processes, and governance models that enable inner-source contribution at scale while maintaining enterprise quality and security requirements.
• Establishes what qualifies as a skill, an agent, and an MCP configuration at the enterprise level.
• Acts as the senior technical resource mentoring engineers across the organization in AI-native engineering practices — including agentic coding patterns, context engineering, prompt-to-code workflows, and AI-assisted testing — raising the floor of capability so teams become self-sustaining without ongoing coaching dependency.
• Researches, evaluates, and recommends AI engineering tools, frameworks, and infrastructure (e.g., eval platforms, agent orchestration systems, environment provisioning automation) aligned with CNA's strategic direction.
• Leads build-vs-buy analysis for platform capabilities such as CI/CD tooling, sandbox provisioning, and LLM evaluation infrastructure.
• Pairs closely with Architecture, Security, Cloud Engineering, and Data teams to ensure the AI engineering platform integrates with enterprise infrastructure (GCP/GKE, GitHub, JFrog Artifactory), meets regulatory and compliance requirements (AI model tracking, Sox controls), and scales to support hundreds of engineers and AI pod teams across all portfolios.

Skills, Knowledge & Abilities

• Expert knowledge of AI-native software engineering practices including agentic coding workflows (Claude Code, Cursor, GitHub Copilot), prompt and context engineering, multi-agent orchestration, MCP protocol, and skill/agent authoring patterns.
• Deep understanding of the modern software delivery lifecycle with specific expertise in how AI transforms each phase — from AI-assisted requirements and design through agentic code generation, automated testing, AI-augmented code review, and continuous deployment.
• Expert-level proficiency in building and operating CI/CD platforms (GitHub Actions or equivalent), infrastructure-as-code (Terraform), container orchestration (GKE/Kubernetes), and cloud platforms (GCP), with the ability to design pipelines that enforce quality and security gates without creating delivery bottlenecks.
• Strong knowledge of application security engineering including supply chain security, artifact management and curation, static/dynamic analysis, secret management, and the specific attack vectors introduced by AI-generated code (dependency hallucination, model drift, prompt injection).
• Demonstrated ability to design developer platforms and tooling that serve hundreds of engineers at varying skill levels — balancing power-user capability with guardrails that prevent misuse and maintain code quality at scale.
• Proven ability to evaluate and integrate emerging AI technologies rapidly, with the judgment to distinguish between hype and production-ready capability.
• Comfortable operating in a fast-moving domain where the tooling landscape changes weekly.
• Excellent communication skills with the ability to translate complex AI engineering concepts for both technical and non-technical audiences.
• Able to influence engineering culture and drive adoption of new practices across a large, diverse organization including internal teams and managed service providers.
• Strong analytical and problem-solving skills with an outcomes-oriented mindset — focused on measurable improvements in delivery speed, code quality, and engineering productivity rather than tooling adoption metrics.

Education & Experience

• Bachelor's Degree with Master's preferred in Computer Science, AI/ML, or related discipline, or equivalent work experience.
• Minimum of 9 years of solid, diverse work experience in software engineering with a minimum of 6 years in application development, including significant recent experience (2+ years) building or operating AI-augmented development tools, agentic systems, or developer platforms.
• Demonstrated hands-on experience with LLM-based engineering tools (Claude Code, Cursor, GitHub Copilot, or equivalent) in production engineering workflows, not just experimental use.
• Experience designing and scaling inner-source or platform engineering programs across large engineering organizations preferred.

Benefits

CNA offers a comprehensive and competitive benefits package to help our employees – and their family members – achieve their physical, financial, emotional and social wellbeing goals. For a detailed look at CNA’s benefits, please visit cnabenefits.com.