Product Manager with 9+ years building and scaling products at the intersection of loyalty, rewards, and marketing technology for Southeast Asia's leading travel platform. Known for leading cross-functional teams from early ideation through global rollout — translating business strategy into measurable outcomes across engineering, design, data science, and commercial functions. Combines technical depth in API-first systems, data-driven experimentation, and AI/ML with sharp business instincts honed across rapidly evolving consumer markets. An M.S. in Information Systems Management from Carnegie Mellon University deepened this foundation — bringing structured frameworks in AI strategy, technology policy, and organizational design to complement nearly a decade of hands-on product delivery.
Featured ↓Grew the rewards program from a single-market product to a multi-country platform — driving 300% MoM user growth, 80% higher redemptions, and unlocking six-figure partnership contracts.
Traveloka's loyalty and rewards program had strong initial adoption but plateauing engagement, low redemption rates, and no multi-market presence. The earn-and-burn model was confusing across tiers, and the product had accumulated operational debt — reconciliation alone took 2 weeks every month.
I owned the loyalty and rewards product end-to-end across three role levels (PM I → PM II → PM III), managing cross-functional teams of up to 3 PAs, 10 engineers, and 4 designers. I drove the program from a single-market feature to a multi-country platform and led partnership negotiations for co-branded products.
300% MoM growth in Monthly Transacting Users across 6 markets. 80% boost in rewards redemption rates. Hundreds of thousands of USD unlocked in partnership revenue via API-first infrastructure. Millions in annual financial penalties prevented through reconciliation automation.
Loyalty is a platform, not a feature. The biggest unlock was treating it as infrastructure — building the API layer first — so that business development, marketing, and finance could all build on top of it without coming back to engineering for every new use case.
By 2019, Traveloka's paid marketing spend was underperforming — fragmented attribution, broad audience targeting, and no unified view of spend efficiency across 6 markets. When COVID-19 hit in 2020, the problem intensified: the company was sitting on a sprawling 20+ tool martech stack with redundant capabilities and contracts negotiated piecemeal across countries.
Starting as a Senior Associate focused on ROAS optimization, I moved into the PM I role to lead the full infrastructure rebuild. I owned marketing technology end-to-end: attribution modeling, tag management, mobile SDK, vendor consolidation, and the budget renegotiations during COVID.
50% improvement in ROAS across all paid channels. 90% reduction in marketing infrastructure costs. 12 months of contract value saved through COVID renegotiations. The rebuilt stack became the foundation for all subsequent marketing technology work at Traveloka.
Crisis creates the political will to do what should have been done already. The COVID budget freeze forced consolidation decisions that had been deferred for years. The hardest part wasn't the technical rebuild — it was getting channel managers to trust the new attribution numbers and change how they allocated spend.
For my capstone consulting project at CMU's Heinz College, I partnered with a global consumer goods company facing a recurring problem: no systematic way to evaluate ROI on significant capital investments or make rationalization decisions across their enterprise tooling portfolio at scale. Average mold investments ran ~$400K, with no consistent framework for lifecycle cost analysis.
Project lead and primary analyst. I scoped the engagement, designed the analytical framework, built the forecasting and cost model, and presented recommendations to senior client stakeholders.
Delivered a reusable TCO framework, forecasting model, and a prioritized rationalization roadmap. The project was recognized as a standout engagement at Heinz College.
Enterprise cost analysis is more political than technical. The model was the easy part — the hard part was getting stakeholders to agree on what "cost" means when every team defines it differently. You need a shared vocabulary before you can build a shared framework.
A leading vision care provider needed a data-driven approach to optimize eyewear frame assortment across hundreds of store locations. The challenge: which frames to stock where, and what actually drives sales volume at the product level?
Led the data analysis and modeling work for a five-person team. I drove the ML approach — selecting algorithms, running experiments, and synthesizing findings into actionable product and inventory recommendations.
Delivered a location-specific inventory strategy backed by ML evidence, with clear prioritization of which product attributes to optimize for different store segments. Recommendations covered assortment breadth, price tiering (budget / mid-range / premium), and service quality improvements surfaced from review sentiment.
The most valuable insight wasn't the model output — it was discovering how much geographic context shapes product preferences. A blanket national assortment strategy leaves significant sales on the table when local demand patterns are this distinct.