Principal Architect, AI-Native Platform Transformation
$229,000–$360,000 year
HybridMcLean, Virginia, United States
Job Summary
Define the Enterprise AI Reference Architecture, establishing standardized agent runtime patterns, memory structures, and interoperability protocols for multi-agent coordination. Set standards for model abstraction, observability, and evaluation to ensure non-deterministic systems remain cost-flexible and traceable. Draw strategic boundaries between centralized platform capabilities and product-owned innovation, leading build-vs-buy decisions to prevent vendor lock-in. Serve as the primary architectural liaison across Product, Engineering, Data, and Security, running review processes to dismantle fragmented AI sprawl through shared services and influence. Establish operational governance for AI lifecycle management, auditability, and deterministic fallback strategies while ensuring compliance with enterprise and government-cloud requirements.
Required Qualifications
- 10+ years of software engineering experience designing and operating large-scale distributed systems and platforms
- deep expertise in backend systems, cloud-native infrastructure, and platform engineering
- Demonstrated, hands-on experience building AI/ML infrastructure, agent orchestration systems, or developer platforms in production
- Demonstrated experience evolving legacy enterprise architectures toward modern, AI-centric or autonomous operational models
- Demonstrated working fluency with the modern agentic stack: LLM serving and routing, agent frameworks and SDKs, tool-integration protocols (MCP or comparable), evaluation infrastructure, and context/memory architectures
- Demonstrated ability to lead complex cross-functional technical initiatives and to drive adoption of architectural standards through influence, clarity, and credibility rather than authority alone
- Demonstrated experience authoring Architecture Decision Records (ADRs), reference architectures, and executive narratives for systems impacting engineering teams
- demonstrated ability to present technical trade-offs to VP-level or C-level stakeholders
- Candidates based in the Tysons vicinity
- Hybrid, 3 days per week onsite
Desired Qualifications
- Deep expertise operationalizing LLMs, multi-agent frameworks, and autonomous workflow paradigms at enterprise scale (LLMOps/AgentOps)
- Knowledge of AI safety, policy enforcement, and responsible AI operational practices, particularly in compliance-sensitive or regulated environments
- Experience with multi-tenant SaaS platform architecture and the particular challenges of per-customer configuration, data isolation, and schema variability
- Track record establishing architecture governance functions (review boards, ADR practices, golden paths) that teams experience as enabling rather than obstructing
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.