Principal AI Architect

Overview
The Principal AI Architect is a senior technical leadership role responsible for defining and governing enterprise-wide AI, GenAI, and MLOps architecture. This role sets the long-term vision, reference architectures, and reusable patterns for scalable, secure, and responsible AI adoption across the organization. This role combines strong hands-on engineering with strategic innovation to design, prototype, and deliver intelligent, data-driven solutions that power analytics, machine learning, and next-generation AI applications.

Responsibilities
Enterprise AI & GenAI Architecture
Define and own enterprise AI and GenAI reference architectures, including LLM platforms, RAG patterns, agentic systems, and multimodal solutions.
Establish standardized architectural patterns for model serving, prompt management, orchestration, tool use, and agent frameworks.
Lead architecture decisions for buy vs. build, model selection, hosting strategies, and vendor integrations.
Ensure AI architectures align with enterprise standards, cloud strategy, security, and governance.
MLOps & AI Platform Engineering

• Prototype and operationalize advanced AI solutions, including GenAI and LLM-based systems.
• Architect end-to-end MLOps capabilities, including model lifecycle management, CI/CD for ML, feature stores, model monitoring, and drift detection.
• Define enterprise patterns for training, fine-tuning, deployment, and observability of ML and GenAI workloads.
• Guide teams on productionizing PoCs into scalable, resilient, and supportable AI systems.
• Partner with platform teams to evolve a shared enterprise AI platform.

AWS-Centric AI Architecture
Design AI and GenAI solutions using AWS-native services, including (but not limited to):
Amazon Bedrock, SageMaker , Lambda, ECS/EKS
S3, DynamoDB, Aurora, OpenSearch
IAM, KMS, VPC, CloudWatch
Define cost, performance, and scalability guardrails for AI workloads on AWS.
Ensure architectures follow Well-Architected Framework principles.
Governance, Risk, and Responsible AI
Partner with security, legal, and compliance teams to define AI governance, guardrails, and controls.
Embed responsible AI principles, data privacy, and explainability into enterprise designs.
Establish standards for model access, auditability, and risk management.

Technical Leadership & Influence

• Act as a principal-level advisor to senior technology and business leaders .
• Champion hands-on experimentation and rapid solution delivery while maintaining technical excellence.
• Mentor architects and senior engineers on AI architecture and MLOps best practices.
• Drive alignment across teams by publishing reference architectures, design standards, and decision frameworks.
• Represent the organization in architecture forums, reviews, and strategic initiatives.

Qualifications

• • Bachelors or Masters degree in Computer Science , Engineering, or a related field
    • 1 0 + years of experience in enterprise architecture, data platforms, or distributed systems.
    • Deep expertise in AI/ML architecture, including GenAI and LLM-based systems.

      Looking to get Placed? Try our Placement Guarantee Plan

      Get Hired
    • Strong experience designing MLOps platforms and enterprise AI foundations.
    • Proven experience architecting solutions on AWS.
    • Strong understanding of cloud security, networking, and governance.
    • Proficiency in Python, SQL, and modern data frameworks (e.g., Databricks, Airflow, Snowflake, Vertex AI).

Preferred:

• Experience with agentic AI design patterns, including tool-use orchestration, autonomous workflow agents, or AI copilots.

• Proficiency in API design, microservices, and containerization (Docker, Kubernetes).

• Demonstrated ability to rapidly prototype new AI concepts and transition successful PoCs into production-grade systems.

Benefits are an integral part of total rewards and First Citizens Bank is committed to providing a competitive, thoughtfully designed and quality benefits program to meet the needs of our associates. More information can be found at ;br>


$descr2
$descr3