AI Application Architect

About the role

As a key member of Philips’ R&D team in Beijing, the AI Application Architect will design, develop, and maintain the AI R&D QMS Harness platform that supports controlled AI workflows for medical-device research. The role focuses on building production-grade AI services, enterprise integrations, and governed workflow agents while ensuring compliance with quality-system controls.

Key responsibilities

• Develop AI/LLM application services, including prompt pipelines, retrieval-augmented generation (RAG) services, workflow orchestration, evidence generation, and reviewer-facing outputs.
• Build and maintain read-only integrations with enterprise systems such as DOORS, ALM, ClearQuest, Windchill, SOP/document repositories, and DHF repositories.
• Implement AI agents for requirement-quality checks, traceability gap detection, DHF completeness review, defect impact analysis, compliance checks, and draft evidence generation.
• Design audit logging, model routing, access control, project isolation, and evidence packaging to enable controlled and traceable AI usage.
• Support operational qualification/performance qualification (OQ/PQ) by creating test harnesses, representative datasets, prompt regression tests, hallucination checks, and workflow validation evidence.

Required profile

• 3+ years of experience in software engineering, AI application development, LLM integration, workflow automation, or enterprise systems integration.
• Strong programming experience in Python and practical experience with APIs, backend services, and structured data processing.
• Hands-on experience building RAG pipelines, prompt-based applications, evaluation workflows, or agent-style systems for business or engineering use cases.
• Familiarity with vector search, document chunking, metadata filtering, citation generation, and prompt/version management.
• Experience working with cloud AI platforms, model APIs, or private inference environments in an enterprise setting.

Required skills

• Python
• API development and integration
• Backend services
• Structured data processing
• AI/LLM application development
• Retrieval-augmented generation (RAG) pipelines
• Vector search and document chunking
• Metadata filtering and citation generation
• Cloud AI platforms and model APIs
• Private inference environments
• Logging, testing, reproducibility, access control, and deployment practices