Head of AI

Job Title

Head of AI Philips China

Job Description

Job Title

Head of AI Philips China

Job Description

You own the AI agenda for Philips in China: which use cases we build, on which stack, to what standard, and how fast they reach production. This is a builder's role with an executive seat — you will sit in business reviews with China leadership and in architecture reviews with engineers, and you will be judged on shipped solutions with measured value, not on a portfolio of pilots.

Your solid line runs to the Global Head of the Philips AI Center of Excellence, which gives you the global reference architecture, the AI Hub, the enterprise partnerships, and the Responsible AI framework. Your dual local line runs to a member of the China Management Team, which gives you the business mandate, the priorities, and the access. You will work in close daily partnership with Philips China IT — they own the local infrastructure, applications, and security landscape; you own the AI capability that runs on it. Neither of you builds alone.

You will build and lead a small, senior, multidisciplinary team in China — AI and ML engineers, data scientists, AI solution architects, and a product-minded lead for adoption — deliberately lean, deliberately close to the businesses.

What you'll build and drive

1. Solutions built with the businesses

Work directly with Health Systems, Personal Health, and the commercial, supply chain, service, quality, and R&D functions in China to find, size, and sequence the AI use cases that matter — then build them. Typical territory: field service and installed-base support, sales and marketing personalization for Chinese digital channels, clinical and customer documentation, demand forecasting and supply planning, localization of global products and content, and consumer engagement in the WeChat ecosystem.

Run a visible intake-to-production pipeline: triage, feasibility, build, evaluate, ship, measure. Kill what does not work early and say so.

Co-own outcomes with the business sponsor. Every solution has a named business owner, a baseline, and a value measure agreed before you build.

2. A deliberate dual-stack AI architecture

Global Philips stack: frontier models from OpenAI and Anthropic consumed through the Philips enterprise AI Hub, Microsoft Azure and AWS AI services, Microsoft 365 Copilot at enterprise scale, and the global agentic patterns, evaluation harnesses, and paved-road SDKs from the AI CoE. Your job is to make these usable in China wherever they legally and technically can be.

China-domestic stack: where the global stack cannot reach — for latency, data residency, cost, Chinese-language quality, or channel integration — build on leading domestic alternatives. Expect to work with Alibaba's Qwen family on Alibaba Cloud (Model Studio / Bailian), ByteDance's Doubao on Volcano Engine, DeepSeek, Tencent Hunyuan and Tencent Cloud, Zhipu GLM, Baidu ERNIE and Baidu AI Cloud, Moonshot Kimi, and Huawei Cloud with Pangu models and Ascend silicon where domestic hardware matters. Vendor choice is a decision you will own and defend on evidence.

Channel and workplace integration where Chinese users actually are: WeChat and WeCom official accounts and Mini Programs, DingTalk or Feishu (Lark) workflows, and the local CRM, service, and commerce systems those channels feed.

The layer that matters most: a model and tool abstraction — gateway, routing, prompt and evaluation management, tracing, and cost control — that lets a single use case run on a global or a domestic model without a rewrite. Portability is the architecture principle; single-vendor lock-in in either direction is the failure mode.

3. One team with China IT and the global CoE

Partner with Philips China IT on infrastructure, identity, security, network, and application integration so AI solutions land on the sanctioned local landscape rather than beside it. You are a demanding customer and a co-owner, not a shadow IT function.

Feed China patterns back into the global AI CoE — domestic model evaluations, cost benchmarks, localization learnings, regulatory practice — and pull global assets in rather than rebuilding them. Where a global asset genuinely cannot work in China, document why and build the local equivalent to the same standard.

Represent Philips China in global AI architecture, governance, and vendor decisions so China constraints are designed for upfront, not retrofitted.

4. Data, sovereignty, and compliance by design

Build the China data foundation AI needs — local lakehouse and feature capability, consented and classified data products, and clear separation between what stays in China and what may lawfully leave — in step with Philips' global data and analytics architecture.

Own the AI compliance posture for China in practice, not on paper: PIPL, the Data Security Law, the Cybersecurity Law and MLPS 2.0 grading, CAC filing and registration for generative AI services and applications, AI-generated content labeling under the 2025 labeling measures and GB 45438-2025, the 2026 measures on anthropomorphic interactive services where relevant, and cross-border personal information transfer via the correct pathway — security assessment, standard contract, or the certification route in force since January 2026.

Manage the boundary between general AI features and regulated medical functionality, working with Regulatory Affairs and Quality on anything approaching NMPA scope. Translate Philips' global Responsible AI principles into engineering practice locally: model documentation, bias and safety evaluation, human oversight, incident response, and audit-ready traceability.

5. Fluency, community, and talent

Raise AI fluency across Philips China — business leaders, commercial teams, engineers — through hands-on enablement rather than awareness slides. Make Copilot and the AI Hub genuinely used, and make it obvious what good looks like.

Build the China AI community of practice and grow the local talent pipeline through hiring, mentoring, and partnerships with Chinese universities, startups, and cloud and model vendors.

Define and report the KPI framework — use cases in production, adoption, model performance, AI spend efficiency, and realized business value — to both the China Management Team and the global AI CoE.

Your first 12–18 months

• A prioritized China AI portfolio exists, agreed with the businesses and the China Management Team, with clear value cases and owners.
• Three or more AI solutions are running in production for Chinese users, at least one built on a domestic model stack and at least one on the global Philips stack, with measured business impact.
• The dual-stack architecture and abstraction layer are live and documented, jointly owned with China IT and endorsed by the global AI CoE.
• The China AI compliance operating model is running — use-case triage, data classification, filing and labeling where required, cross-border transfer pathways — and teams experience it as an accelerator, not a queue.
• A lean China AI team is hired and productive, with clear charters and a visible community of practice around it.
• 
  

What you bring

Must-have

• A track record of shipping AI/ML solutions into production in China — real systems with users, SLOs, incidents, and measured outcomes, not strategy decks or pilots.
• Deep working fluency in the modern AI stack: LLMs and generative AI, agentic frameworks and orchestration, RAG and vector search, MLOps/LLMOps, evaluation and observability, and cloud-native architecture.
• Hands-on experience with the Chinese AI ecosystem — domestic foundation models and the clouds that serve them (Alibaba Cloud, Volcano Engine, Tencent Cloud, Baidu AI Cloud, Huawei Cloud) — combined with credible experience of at least one international stack (Azure AI, AWS, OpenAI, Anthropic). We need someone fluent in both, not one or the other.
• Practical command of China's data and AI regulatory regime: PIPL, DSL, CSL and MLPS, generative AI filing, content labeling, and cross-border data transfer mechanisms — enough to design for them, not just cite them.
• Proven delivery inside a large matrixed multinational in China: managing a dual reporting line, partnering with a local IT organization, influencing global functions, and getting decisions made across a headquarters-and-market boundary.
• Experience building and leading small senior technical teams, and the personal credibility to stay hands-on in architecture and code review.
• Professional fluency in Mandarin and English. You will present to Chinese business leaders and to a global CoE in the same week.

Differentiators

• Healthcare, MedTech, or consumer health domain experience in China, including familiarity with NMPA expectations for AI-enabled products.
• Experience integrating AI into the WeChat/WeCom ecosystem, DingTalk, or Feishu at enterprise scale.
• Experience with domestic AI silicon and on-premise or private-cloud model deployment (Huawei Ascend, Cambricon) where data residency demands it.
• AI cost engineering at scale — token economics, cost/quality routing across domestic and international models, capacity strategy.
• Edge and embedded AI on connected devices, or privacy-preserving techniques such as federated learning in production.
• A public technical footprint: publications, patents, open-source contributions, or conference talks in the Chinese or international AI community.
• Typically this profile comes with 12+ years in AI, data, or platform engineering, including several years leading in China — but we care about what you have shipped and scaled, not the exact year count.

Why this role, here

• A real mandate: a dual line to the global AI CoE and to the China Management Team means you have both the technology backing and the business authority to make things happen — and nowhere to hide.
• The best of both ecosystems: Philips' global partnerships with Microsoft, AWS, OpenAI, Anthropic, and NVIDIA, plus the freedom and expectation to build on China's fastest-moving domestic AI stack where it is the better answer.
• Breadth few country AI roles offer: hospitals and clinicians on one side, hundreds of millions of consumer health moments on the other, and a local R&D and manufacturing base in between.
• Responsible AI with teeth: published principles, decades of experience shipping AI in regulated health contexts, and a China compliance model you will help build rather than fight.
• Purpose that survives contact with Monday morning: faster diagnosis in a county hospital, better daily health habits in millions of Chinese homes. Improving lives is the product, not the tagline.
• If you want to build AI that has to work in the real world — under China's constraints, at Philips' scale, for both patients and consumers — this is your moment. Apply, or reach out for a confidential conversation.
• Philips is an equal-opportunity employer. We value diverse teams and are committed to an inclusive recruitment process; accommodations are available on request.

Preferred Skills:
• Statistical Methods

• Entrepreneurship and FDE capability

• Data Harmonization & Processing
• Artificial Intelligence (AI)
• AI Algorithm Development
• DevOps
• Business Acumen
• Data Governance
• Continuous Improvement
• Data Warehousing (DW)
• Strategic Planning
• Regulatory Requirements
• People Management
• Stakeholder Management