We are seeking an experienced Agentic AI Architect to define the enterprise architecture for reusable AI products and agentic platforms. This role owns the technical direction for secure, scalable, and governable solutions built with the same core technology stack used by the Agentic AI engineering team. The Architect will convert business priorities into reference architectures, reusable patterns, and delivery standards that multiple product teams can adopt.
The ideal candidate brings 12 or more years of experience across software, data, cloud, or AI engineering, including at least 3 to 5 years designing GenAI or LLM-based systems. The role requires both enterprise-level architectural judgment and enough hands-on depth to guide implementation, review designs, resolve technical risks, and support delivery teams.
What You Will Own
· Enterprise AI architecture: Define the target architecture, reference patterns, and technology standards for AI-enabled products, services, and automation solutions across the enterprise.
· Reusable agent platform: Create reusable components for agent orchestration, tool integration, state management, prompts, security, evaluation, observability, and human approval so teams do not rebuild common capabilities.
· Agentic solution design: Architect GenAI and agentic AI applications with LangChain and LangGraph, integrating Azure OpenAI and leading LLM models into enterprise workflows.
· RAG and knowledge architecture: Define ingestion, grounding, retrieval, metadata, access-control, and evaluation patterns for RAG solutions using PostgreSQL and PgVector.
· Application and integration architecture: Establish Python, FastAPI, REST API, and microservice patterns for reliable integration with enterprise systems, data sources, tools, and business processes.
· Cloud and delivery architecture: Define deployment patterns and environment standards for AKS, ACA, and SWA, with automated delivery through GitHub Actions and ArgoCD.
· Nonfunctional architecture: Set requirements and design decisions for security, privacy, scalability, resilience, latency, cost, maintainability, and operational support.
· Design authority: Lead architecture reviews, document trade-offs and decisions, resolve cross-system dependencies, and ensure implementations remain aligned with approved enterprise patterns.
· Technical leadership: Guide and mentor engineering and QE teams, support complex troubleshooting, and maintain a practical connection between architecture and implementation.
· Stakeholder partnership: Partner with product leaders, business stakeholders, security, compliance, data teams, and delivery leadership to prioritize use cases and maintain an executable AI roadmap.
Architecture Leadership and Governance
· Reference architectures: Define repeatable patterns for single-agent, multi-step, tool-using, and human-in-the-loop workflows, including clear boundaries between deterministic services and model-driven decisions.
· Guardrails and controls: Establish standards for prompt and tool safety, access control, data protection, auditability, responsible AI, and escalation of high-impact actions.
· Lifecycle management: Define how AI solutions are designed, reviewed, tested, deployed, monitored, versioned, improved, and retired across environments.
· Evaluation and quality gates: Set measurable acceptance criteria for retrieval quality, answer quality, tool execution, safety, latency, cost, reliability, and regression testing before release.
· Architecture assets: Maintain reference diagrams, decision records, interface contracts, reusable templates, implementation guidance, and approved technology patterns for delivery teams.
· Executive and client guidance: Explain architectural options, risks, dependencies, investment choices, and adoption roadmaps clearly to senior technical and business stakeholders.
Technology Environment
Area | Technologies |
AI and agent frameworks | GenAI, Agentic AI, LangChain, LangGraph |
LLM and retrieval | Azure OpenAI Services, LLM Models, Prompt Engineering, RAG Architecture |
Application development | Python Development, FastAPI, REST APIs, Microservices |
Data and vector search | PostgreSQL, PgVector |
Delivery and infrastructure | GitHub Actions, ArgoCD, Azure Kubernetes Service (AKS), Azure Container Apps (ACA), Azure Static Web Apps (SWA) |
What You Bring
· Architecture experience: 12 or more years of experience in software, data, cloud, or AI engineering, including 3 to 5 or more years with GenAI, LLM, or advanced AI solutions.
· Enterprise design leadership: Proven ownership of end-to-end solution architecture, reference architectures, reusable platforms, and architecture decisions across multiple teams or products.
· Hands-on technical depth: Strong working knowledge of LangChain, LangGraph, Azure OpenAI, Python, FastAPI, REST APIs, PostgreSQL, and PgVector.
· Cloud-native delivery: Experience designing and governing containerized workloads and CI/CD delivery using GitHub Actions, ArgoCD, AKS, ACA, and SWA.
· Enterprise integration: Strong background in APIs, microservices, distributed systems, data integration, authentication, authorization, and secure enterprise connectivity.
· AI quality and governance: Ability to define evaluation strategies, guardrails, monitoring, lifecycle controls, and production-readiness standards for LLM and agentic systems.
· Architectural judgment: Ability to compare options, explain trade-offs, manage technical debt, and make decisions that balance delivery speed with security, reliability, cost, and long-term reuse.
· Leadership and communication: Strong written and verbal communication, with the ability to influence senior stakeholders, lead design reviews, and mentor architects and engineers.
· Execution focus: Demonstrated ownership, adaptability, and the ability to turn enterprise architecture into practical standards and working solutions.
· Continuous learning: Strong curiosity for emerging AI capabilities and the discipline to evaluate new approaches against enterprise requirements before adoption.
