Role mission
This role ensures that the company chooses the right AI opportunities and turns the most important business problems into AI products that are useful, measurable, safe and scalable.
This is not a traditional PMO role and not a pure engineering role. The position combines portfolio judgment, product leadership and forward-deployed delivery. The successful candidate will work directly with business teams, understand how work is actually performed, build or shape working prototypes, and lead selected use cases from discovery through pilot and scale decision.
Why this role exists
The company has growing employee adoption, business demand and access to AI platforms. The next challenge is converting that energy into sustained business outcomes rather than disconnected assistants, demonstrations or one-off experiments.
This role connects business priorities, domain expertise, AI technology, data, risk and organizational adoption. It provides the discipline to invest in the right opportunities, stop weak projects early, and turn validated solutions into reusable organizational capabilities.
Key responsibilities
1. Manage the AI opportunity portfolio
- Maintain a transparent portfolio of AI opportunities, active initiatives, owners, lifecycle stages, evidence and resource needs.
- Translate informal ideas into clear problem statements, target users, business outcomes and decision requirements.
- Distinguish personal productivity opportunities from team tools, business products and end-to-end workflow transformation.
- Recommend which opportunities should enter discovery, be combined, paused, redesigned, scaled or stopped.
- Prepare monthly portfolio reviews and executive decision materials with clear facts, assumptions, risks and next-stage recommendations.
- Identify reusable patterns across brands, functions and métiers to prevent duplicated investment.
2. Lead business and workflow discovery
- Embed with business teams to understand real workflows, decisions, handoffs, exceptions and quality standards.
- Work with Business Owners and Domain Experts to define current-state baselines and measurable target outcomes.
- Identify which activities should be assisted, augmented, automated or redesigned with AI.
- Define the human accountability boundary: what AI may recommend or execute, what requires review, and when the workflow must stop or return to a person.
- Select representative real cases for validation rather than relying on curated demonstrations.
3. Design and validate AI products
- Convert business problems and expert judgment into product requirements, AI workflows, evaluation criteria and rollout hypotheses.
- Build or directly shape lightweight prototypes using approved models, APIs, workflow tools, retrieval, data processing or code as appropriate.
- Partner with AI Builders, Data, IT and platform teams on architecture, integration and production-readiness decisions.
- Establish evaluation sets covering task quality, business usefulness, risk, failure modes and exception handling.
- Compare AI-assisted workflows with the current process using real work, clear denominators and realistic human review effort.
4. Drive pilots and adoption
- Form and coordinate cross-functional Use Case Squads with Business Owners, Domain Experts, AI Builders, Data or Process Owners, Change Champions and Risk partners.
- Define pilot scope, roles, milestones, success measures, rollback paths and stop criteria.
- Manage the product backlog and make trade-offs among value, speed, quality, cost and risk.
- Observe actual user behavior, including old-process fallback, manual cleanup and reasons for non-adoption.
- Help managers update working expectations, review practices and team routines when a pilot changes how work is performed.
5. Prepare scale, transfer and reuse
- Build the evidence and business case required for Scale, Redesign or Stop decisions.
- Confirm that value estimates include build cost, platform cost, model usage, human review, exception handling, support and change effort.
- Define the long-term Business Owner, Product or Run Owner, support model, monitoring and continuous-improvement responsibilities.
- Package validated workflows, prompts, evaluation sets, decision rules, playbooks and implementation patterns as reusable assets.
- Feed validated work changes into capability development, AI Champion identification and future job or role planning.
Required experience and capabilities
- 7+ years in product management, digital transformation, enterprise technology delivery, applied AI, management consulting with proven hands-on case delivery or a comparable field.
- Demonstrated ownership of ambiguous, cross-functional initiatives from problem discovery through operational adoption.
- Strong business and workflow discovery skills; able to identify real decision points, hidden exception work and measurable value.
- Hands-on experience with generative AI products, including prompt and workflow design, evaluation and responsible-use controls.
- Sufficient technical fluency to prototype and to challenge system design involving APIs, enterprise data, retrieval, agents, permissions, logging and human-in-the-loop controls.
- Strong product judgment and the willingness to reduce scope or stop projects when evidence is weak.
- Ability to communicate with senior business leaders while working credibly with engineers, data teams and frontline users.
- Experience defining baselines, KPIs, adoption measures and realistic value cases.
- Strong written and verbal communication in English; professional Chinese capability is highly valuable for China business engagement.
Preferred experience
- Experience in beauty, consumer goods, retail, e-commerce, CRM, marketing, supply chain or another judgment-intensive consumer business.
- Experience as an AI Product Lead, Forward Deployed Engineer or digital transformation lead.
- Practical capability in Python, SQL, low-code workflow automation or comparable prototyping tools.
- Experience deploying AI into enterprise workflows with data, security, privacy, legal or regulatory constraints.
- Experience designing evaluation sets, monitoring, exception handling and production adoption mechanisms.
- Experience scaling a successful local solution into a reusable platform, product or operating standard.
