Senior Product Builder - Consumer AI

The Opportunity

ShopBack serves 45M+ shoppers across Asia with cashback, deals, and discovery. We're now building the layer on top: AI-native shopping experiences that personalise, guide, and delight at scale.

This role is for someone who closes the gap between product thinking and design craft - and ships AI-powered consumer experiences that people genuinely use. Not a PM with design opinions. Not a designer with product instincts. A builder who operates at the intersection and holds the bar across both.

You'll work on consumer-facing AI surfaces - search, discovery, personalised recommendations - in a team that has moved past "add AI to the product" and is building AI-first from the ground up.

What You'll Own

Frame the problem, not just the solution.

• You take a messy space - shifting user behaviour, an emerging AI capability, a business bet - and sharpen it into a crisp problem statement with a clear success condition.

Set the craft bar.

• You define what great looks like for your surface. You review work, call out the delta between good and great, and hold that standard under deadline pressure.

Prototype to think.

• You use AI-assisted tools to build working prototypes before specs are written. You don't wait for the design file to be finished before forming a point of view.

Lead the product strategy for your domain.

• You own the outcome, not just the output. You connect metric movements to specific design decisions and change direction when the data doesn't support the hypothesis.

Build the eval layer for your AI surfaces.

• When an AI feature ships, you own the quality bar - structured test cases, scoring criteria, and failure mode coverage. AI output is unverified until tested. "It looks right" is not a quality check.

Lift the team around you.

• You actively shape how other builders approach problems - through review, patterns, and shared standards. Leadership is a day-1 expectation at this level.

Model AI-native practice for the team.

• At this level, your job isn't just to use AI fluently - it's to help the people around you understand where it creates real leverage in their specific work. Not sharing tools. Demonstrating deliberate practice.

Maintain the design system as a shared knowledge layer.

• Token architecture, component docs, and pattern rationale kept in formats that engineers and AI tools can consume. Not just Figma files.

What We're Looking For

Craft and product, not one or the other

• Strong design eye with the ability to prototype consumer experiences end-to-end - from rough layout to precise interaction detail
• Can spec requirements in forms that directly unblock engineers: verbal description, production URL reference, component vocabulary, or working prototype
• Work has very few visual issues; holds the line on localisation, edge cases, and interaction precision - not just the happy path

Genuine fluency with consumer AI products

• Has shipped AI-powered features that real users relied on - search, recommendations, generative UI, conversational interfaces, or similar
• Understands where AI adds real value and where it adds noise; can articulate the design principles that distinguish the two

AI as a primary co-worker, not an occasional shortcut

• You direct AI across your daily workflow - copy iteration, UI pattern generation, prototype production, research synthesis, localisation review - and validate outputs against your own judgment. You run materially faster than a conventional workflow without offloading the judgment layer. This is a baseline expectation for this role, not a differentiator.

Seniority that scales beyond yourself

• Independently strategic - doesn't need the problem handed to them
• At least one team around them is doing better work because of their involvement
• Track record of making the right call in ambiguous, cross-functional situations

The building/selling test

• Can point to a consumer metric that moved because of a design decision they owned
• Can show work that others built on - a component, a pattern, a framing - not just work that shipped

How We Work

• We have moved past "using AI tools." The team operates in an AI-native workflow: prototypes generated before specs, design systems maintained as machine-readable knowledge layers, eval sets built for AI-assisted surfaces, and no-canvas mode the default for incremental changes.
• AI output on this team is adversarially evaluated, not accepted. Eval sets are shared assets, not personal checklists.
• You will be expected to adopt, model, and evolve these practices - not just follow them.