Staff AI Architect
$191,200–$239,000 year
HybridBoston, Massachusetts, United States
Job Summary
Define the AI reference architecture for agentic systems, establishing patterns for orchestration, tool use, and observability while enforcing standards through tooling rather than approval boards. Design and code the internal AI platform, including model routing, agent runtimes, and durable orchestration for long-running workflows, alongside the tool layer built on the Model Context Protocol. Partner with functional leaders in finance, sales, and support to re-architect business processes into AI-native workflows, ensuring integration with enterprise systems like Workday and Salesforce without replacing native authorization. Set the technical bar for governance and safety by embedding capability boundaries, audit trails, and cost-per-task economics directly into the architecture. Mentor engineers across distributed teams and serve as the escalation point for complex design decisions, ensuring interfaces remain stable for future vendor swaps.
Required Qualifications
- Substantial experience as a software, solution, or enterprise architect (typically 10+ years)
- Several years owning architecture above a single project or team
- Experience transforming a project (including what went wrong and what was changed)
- Recent experience building LLM and agentic systems that ran in production
- Experience with agent orchestration (LangGraph, CrewAI, AutoGen, Semantic Kernel, OpenAI Agents SDK, or equivalents)
- Experience with Model Context Protocol (MCP) tooling
- Experience with retrieval and vector stores
- LLMOps discipline including evaluation-first development, prompt and agent versioning, regression testing, and observability for non-deterministic outputs
- Experience with cost attribution
- Depth in at least one production language (Python, Go, TypeScript, or Java)
- Experience with cloud-native infrastructure (Kubernetes, serverless, APIs, event-driven patterns, observability)
- Experience architecting on and integrating Workday, Salesforce, NetSuite, Greenhouse, ServiceNow, or similar systems
- Knowledge of their data models, extensibility limits, native agent layers, and permissioning models
- Experience with API and event-driven design
- Experience with workflow and iPaaS platforms (Workato, MuleSoft, Boomi, n8n, or equivalents)
- Experience with cloud data platforms
- A clear position on securing autonomous systems (agents as first-class principals, short-lived machine identity, vault-backed scoped secrets, delegation with preserved provenance, default-deny tool access, prompt-injection defense, audit trails)
- Familiarity with the OWASP Agentic AI risk landscape
- Fluency with the NIST AI RMF, ISO/IEC 42001, and the EU AI Act
- Excellent written and verbal communication
- The ability to influence engineers, executives, and non-engineering stakeholders without authority
- Effective collaboration across time zones, including close partnership with engineering teams in India
- Comfort in a hybrid environment near Boston/Cambridge
Desired Qualifications
- Experience deploying AI developer tooling at scale to internal users (Cursor, Claude Code, GitHub Copilot, or equivalent enterprise rollouts)
- Experience re-engineering finance, people, GTM, or support processes in enterprise systems, including SOX-relevant or otherwise audited workflows
- Familiarity with emerging agent interoperability and identity work (A2A, agent registries, SPIFFE/SPIRE, and the MCP authorization spec)
- Experience with agent evaluation and observability tooling (LangFuse, Arize, Braintrust, LangSmith, OpenTelemetry-based tracing, or equivalents)
- Knowledge of knowledge graphs, semantic layers, or ontology modeling applied to enterprise retrieval, including GraphRAG patterns
- Enterprise architecture practice experience (reference architectures, ADRs, C4 modeling, portfolio rationalization) or a framework background such as TOGAF
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.