Principal AI Architect

Important: We are specifically seeking candidates with AI Development Lifecycle (AI DLC) experience, and this requirement is mandatory for the role.

Job Title: Principal AI Architect

Job Location: Edina, MN(Hybrid 3 Days/Week)

Summary

The Associate Principal, AI Optimization & Delivery Solution Architecture serves as the technical cornerstone of Client's AI Optimization & Delivery organization. As a principal individual contributor with a greenfield mandate, this role acts as the primary technical design authority for AI Optimization & Delivery and is responsible for establishing the architecture, standards, governance, and operating model for enterprise-scale AI adoption.

This position partners with AI Optimization & Delivery leadership, Enterprise Architecture, Data, Security, Clinical, and Engineering teams to ensure Client's AI ecosystem integrates cohesively with enterprise platforms and supports adoption across distributed engineering teams.

AI DLC Requirement (Mandatory)

A strong background in the AI Development Lifecycle (AI DLC) is mandatory for this role. The successful candidate must have demonstrated experience leading and governing the complete AI lifecycle, including:

• AI strategy and use case identification
• Data acquisition and preparation
• Feature engineering and model development
• LLM and Agentic AI solution design
• Evaluation and validation frameworks
• Deployment and operationalization
• Monitoring, observability, and governance
• Model retraining and continuous optimization

The ideal candidate must also have hands-on experience delivering both:

• Greenfield AI initiatives involving the design and implementation of AI platforms and capabilities from the ground up.
• Brownfield AI programs focused on modernizing, integrating, scaling, and governing AI capabilities within existing enterprise environments.

The immediate focus is to bring architectural clarity to existing AI pilot initiatives by designing blueprints for an agentic component layer, establishing technical exit criteria for sandbox-to-production transitions, implementing AI DLC governance standards, and laying the foundation for a managed AI Internal Developer Platform (IDP).

Over time, this role will architect an ecosystem that enables enterprise engineering teams to deliver AI-powered solutions at scale, accelerate delivery velocity, minimize technical debt, and ensure responsible and governed AI adoption.

Responsibilities

Technical Architecture and Standards

• Establish enterprise architecture standards and design patterns for LLM-enabled workflows, agentic systems, multi-agent orchestration, and AI-powered applications.
• Define and govern architecture standards across the complete AI Development Lifecycle (AI DLC), ensuring consistency from experimentation through production deployment and ongoing optimization.
• Design hybrid system architectures that balance LLM-driven reasoning with deterministic software components for compliance-sensitive and high-confidence business processes.
• Architect the platform component layer, including:
  • Shared AI libraries
  • Reusable agent templates
  • Retrieval-Augmented Generation (RAG) pipelines
  • Vector databases
  • Hybrid search architectures
  • Prompt management frameworks
• Establish reusable AI patterns and contribution models that engineering teams can adopt, extend, and maintain.
• Define the enterprise AI Bill of Materials (AIBOM), including:
  • Model provenance tracking
  • Third-party model risk assessments
  • Version governance
  • Licensing and compliance controls
• Establish observability standards supporting telemetry, reasoning traceability, monitoring, explainability, evaluation, and auditability.
• Design intelligent routing architectures incorporating:
  • Semantic caching
  • Model fallback strategies
  • Cost-aware orchestration
  • Workload distribution between SLMs and frontier LLMs
• Define architectural standards for model lifecycle management, including evaluation, deployment, monitoring, drift detection, retraining, and retirement.

AI DLC Governance and Delivery Architecture

• Define enterprise AI DLC frameworks, governance controls, lifecycle stages, and quality gates.
• Establish architecture review processes ensuring solutions comply with AI DLC standards before progressing between lifecycle stages.
• Define model governance standards covering:
  • Explainability
  • Reproducibility
  • Auditability
  • Risk management
  • Responsible AI principles
• Create engineering guardrails that promote scalable and repeatable AI delivery across the enterprise.
• Establish AI readiness and production certification standards prior to deployment.

Platform Architecture and Governance

• Define technical exit criteria and architectural standards for sandbox-to-production graduation.
• Partner with AI Operations (AIOps) and delivery teams to operationalize successful pilots into governed enterprise platforms.
• Architect enterprise data models supporting:
  • Agent registries
  • Memory architectures
  • Context management frameworks
  • Tool connectivity models
  • Knowledge graph integration
• Lead technical evaluations of AI technologies, platforms, frameworks, and protocols while documenting architectural rationale and recommendations.
• Architect privacy, security, and governance controls tailored for healthcare data environments.
• Ensure AI, RAG, memory, and context-management architectures comply with:
  • HIPAA requirements
  • PHI handling standards
  • PII protection practices
  • Clinical data-sharing agreements
• Establish governance controls for enterprise-scale AI adoption across both Greenfield and Brownfield initiatives.

Platform Evolution and Maturity

• Define year-one and multi-year technical milestones for Client's AI maturity roadmap.
• Establish AI DLC metrics and operational KPIs that measure platform adoption, solution quality, governance compliance, and business outcomes.
• Maintain awareness of emerging AI frameworks, engineering practices, and architecture patterns while translating relevant innovations into platform recommendations.
• Establish evaluation methodologies including:
  • LLM-as-Judge calibration
  • Prompt testing
  • Regression validation
  • Benchmarking
  • Hallucination testing
  • Safety validation
• Integrate evaluation and validation capabilities throughout engineering and deployment pipelines.
• Partner with pilot teams to identify reusable patterns and continuously strengthen platform capabilities.
• Design architectures for:
  • Rate limiting
  • Token throttling
  • Capacity planning
  • Model cost allocation
  • FinOps reporting
• Deliver instrumentation enabling workflow-level AI cost visibility, governance, and optimization.
• Drive continuous improvements based on lessons learned from both Greenfield initiatives and Brownfield modernization programs.

Leadership and Delivery Enablement

• Serve as the technical design authority for AI delivery pods and own solution blueprints aligned with AI DLC governance requirements.
• Eliminate architectural ambiguity early in the lifecycle to accelerate delivery while preventing technical debt.
• Lead architectural governance reviews for AI initiatives throughout their lifecycle.
• Partner with Enterprise Architecture to align AI platforms with broader technology strategies.
• Represent and maintain Client's AI architectural vision while coordinating with enterprise-wide architecture functions.
• Collaborate with Engineering and Operations teams to automate deployments and embed governance controls through policy-as-code frameworks.
• Mentor engineering teams on AI architecture, AI DLC best practices, platform standards, and responsible AI delivery methodologies.
• Partner with Security, Clinical, Product, Data, and Business stakeholders to translate regulatory and operational requirements into scalable platform capabilities.
• Present architecture roadmaps, technology evaluations, and recommendations to executive stakeholders.
• Maintain reference architectures, design patterns, standards repositories, and architectural decision records that enable engineering autonomy.

Qualifications

Required

• 12+ years of experience in enterprise software architecture, AI/ML systems, platform architecture, data architecture, or large-scale technology delivery.
• 5+ years of hands-on experience owning and governing the complete AI Development Lifecycle (AI DLC), including model development, deployment, monitoring, evaluation, governance, retraining, and optimization.
• Demonstrated experience delivering both Greenfield AI programs and Brownfield AI modernization initiatives within enterprise environments.
• 2+ years of hands-on experience delivering production-scale LLM and Agentic AI solutions.
• Experience with:
  • Multi-agent orchestration
  • Context management
  • Tool integration
  • Agent frameworks
• Strong systems-thinking capabilities and the ability to identify architectural risks, failure modes, and downstream impacts.
• Proven ability to make, communicate, and defend enterprise architectural decisions.
• Deep experience designing AI platform infrastructure, including:
  • MLOps
  • LLMOps
  • ModelOps
  • CI/CD pipelines
  • Observability platforms
  • Monitoring frameworks
• Strong understanding of AI model governance and responsible AI principles.
• Experience implementing enterprise AI lifecycle controls and governance processes.
• Strong experience with LLM evaluation methodologies, validation frameworks, and probabilistic system testing.
• Excellent communication, documentation, and stakeholder-management skills.
• Experience operating within regulated environments requiring governance, audit readiness, security controls, and compliance oversight.