AI Architect (Engineering AI & Agentic Systems)

Quick Facts

• Role: AI Architect (Engineering AI & Agentic Systems)

• Work model: Remote

Description

Lead the engineering AI architecture and strategy, evolving a reference architecture integrated with central IT AI services. Design domain-specific engineering capabilities (ontology/knowledge graphs, grounded retrieval, reusable agent tools) and build agent and MCP integrations across CAD/PLM/CAE workflows. Provide technical due diligence on vendor AI solutions and ensure secure, production-ready delivery.

Responsibilities

• Own and evolve the engineering AI reference architecture and integration points with central IT AI services (gateway, model access, identity, agent registry, orchestration, shared observability).

• Architect engineering domain capabilities: engineering ontology/knowledge graphs (when justified), semantic/context services, retrieval, reusable agent tools, and engineering workflow patterns.

• Make build-versus-buy and technical-pattern decisions across LLM agents, deterministic rules, geometric computation, simulation, surrogate models, and hybrid approaches.

• Define traceability between engineering artifacts (requirements, CAD models, parts, assemblies, BOMs, materials, revisions, simulations, test results, and documentation).

• Architect grounded retrieval and engineering RAG/graph retrieval with revision-aware context and fine-grained access controls.

• Lead architecture and hands-on development of multi-step engineering AI agents (tool calling, state, approvals, exception handling, auditable actions).

• Design, build, and sustain engineering MCP servers, API adapters, and integration services for CAD/PLM/CAE and adjacent applications.

• Establish secure authentication/authorization, interface contracts, versioning, CI/CD, evaluation, monitoring, and production support for agents and MCP services.

• Conduct technical due diligence and architecture reviews for vendor solutions; challenge mechanisms, CAD geometry interpretation, simulation validity, interfaces, data handling, security, accuracy, extensibility, and operational support.

• Define technical acceptance criteria, reference architectures, reusable assets, documentation, and handover/ownership boundaries to reduce vendor lock-in.

• Partner on AI security and governance for integrating new and existing vendor solutions.

• Own solution architecture from feasibility/POC to production for engineering AI use cases (e.g., Engineering Admin Assistant, automated design review, simulation workflow automation).

• Translate engineering SME needs into technical designs; lead design reviews, code reviews, and release readiness.

• Hire, mentor, and technically lead agent developers and engineering integration/MCP engineers; scale knowledge/data and AI software engineering support.

Requirements

• Bachelor’s or Master’s degree in computer science, software engineering, engineering, AI, or related discipline; typically 12+ years of relevant software/architecture experience.

• Demonstrated hands-on production delivery of generative AI and agentic AI solutions (LLM tool use, orchestration, retrieval, evaluation, runtime reliability).

• Strong Python/backend development and enterprise API/integration experience; practical MCP/server-tool integration experience or equivalent.

• Meaningful engineering applications knowledge across at least two of CAD, PLM, CAE/simulation, engineering requirements, or NPD workflows.

• Experience with enterprise architecture, cloud, identity and access control, DevSecOps, deployment operations, and technical risk assessment.

• Ability to challenge vendors at implementation depth and lead technical teams across internal stakeholders and partners.

Benefits

First-year outcomes include an approved engineering AI reference architecture, reusable secure engineering context/agent/MCP patterns across use cases, production-ready architecture with measurable quality controls, documented vendor-review standards and technical acceptance checklists, and a sustainable internal engineering AI operations model reducing dependence on single partners.