AI Domain Architect

About the role

PrismHR is hiring an AI Domain Architect, a senior individual contributor who defines the AI standards, reference architectures, approved model catalog and governance for AI across the company. The role owns the domain intelligence layer: the ontology, agent workflow models, fine-tuning strategy and the pattern library that product teams build on. It is a hands-on role where patterns are proven through working proof-of-concepts before becoming standards.

Key responsibilities

• Design and own the canonical ontology as a semantic layer across PrismHR platforms, anchored in the central PEO and HCM platforms, covering co-employment, payroll, benefits, workers' compensation, onboarding and compliance
• Ground reference architectures in that ontology so retrieval, reasoning and generated output stay consistent with the business and its regulations
• Establish governance for ontology evolution and maintain it as a living artifact with a defined review cadence
• Model business processes as agentic workflows, defining autonomy boundaries and embedding safeguards such as validated identity claims, audit logging and content safety
• Own the fine-tuning strategy, dataset curation standards and evaluation harnesses, deciding when to fine-tune, retrieve or prompt
• Build and maintain a reusable library of AI patterns, validating each through a working proof-of-concept and documenting usage, failure modes and evaluation criteria
• Work with product teams from design through go-live, advising on feasibility and risk and acting as escalation point for domain-AI design questions

Required profile

• Eight or more years building production software, including time at staff, principal or architect scope
• Experience designing ontologies, knowledge graphs or canonical domain models that shipped in production
• Hands-on experience with LLM-based or agentic systems, fine-tuning, RAG or model evaluation pipelines
• Ability to move a pattern from concept through proof-of-concept to published standard, and to steer technical direction through influence
• Clear writing, as the ontology, patterns and decision records are part of the output

Required skills

• Microsoft Foundry services, including model deployment, agent development and evaluation and safety tooling
• Azure compute, data, identity and networking
• Preferred: OWL, RDF, SHACL, property graphs, multi-tenant SaaS, SOC 2 or ISO 27001, Microsoft Agent Framework, MCP, Snowflake Cortex AI