About TravelokaTraveloka is Southeast Asia’s leading all-in-one travel platform connecting millions globally with trusted, world-class experiences. With over 140 million app downloads and more than 40 million monthly active users, Traveloka stands as one of the most popular travel apps in the region.

Job Description

Job ID: MJ000325

Traveloka's insurance business sits at the intersection of travel intent and risk. Every time a customer books a flight, hotel, or activity, we have a narrow window to offer the right protection product at the right price — and pricing is one of our biggest levers on both attach rate and margin.

We're hiring a Lead Data Scientist to own that lever. You'll build and run the pricing models behind our insurance cross-sell products: models that serve live, per-request prices across multiple markets, and that let us test pricing strategies rather than guess at them.

This is a hands-on, higher-ownership role. You'll take pricing from problem framing with business stakeholders through modelling and into production — with our ML Platform team providing the serving frameworks and infrastructure, so you're building on established rails rather than from scratch. The centre of gravity is the modelling and the business judgement around it; you just need to be comfortable enough with production systems to ship your own work and keep it healthy.

Responsibilities

• Own insurance pricing end-to-end. Partner with PMs, data analysts, and engineers to translate business problems into pricing models that support dynamic, data-driven strategies — and define the KPIs that prove whether they work.
• Design and train the models. Curate market-specific training data with SQL, run EDA that actually informs feature design, and train, tune, and refresh pricing models on a regular cadence.
• Validate before you ship. Benchmark performance through pricing simulations and backtesting so we understand robustness, revenue impact, and failure modes before anything touches live traffic.
• Ship to production and keep it running. With the support of our Data Engineering, Data Analytics, and ML Platform team, extend our inference serving layer to support experimentation across pricing strategies — revenue maximisation, optimal pricing, confidence thresholds, seasonal discounting — and handle the day-to-day of a live model: deployments, feature freshness, monitoring, drift detection, and root-causing issues when they surface.
• Make the work legible. Communicate approaches, trade-offs, and results to technical and non-technical audiences, and turn model outputs into recommendations leadership can act on.
• Set the direction for pricing science. As the sole owner of this area, you define the roadmap, the methodology, and the standards — and bring stakeholders along with you.

What We're Looking For

Must-haves

• 6+ years of professional experience in data science or applied ML, with a track record of owning ML problems end-to-end.
• Experience with models serving live production traffic, not just batch scoring. You understand what changes when a model has to answer in milliseconds: feature freshness, fallbacks, and what "wrong" looks like at request time.
• Strong SQL and Python, and fluency with the modelling toolkit (gradient boosting, regression, causal or uplift methods as appropriate — plus the judgement to know when the simple model wins).
• Solid grounding in experimentation and causal inference: A/B testing, holdouts, backtesting, and the discipline to distinguish a real lift from a seasonal one.
• Working engineering fundamentals — version control, CI/CD, containerisation, basic cloud and monitoring literacy. Enough to be self-sufficient with platform support.
• Comfort working directly with business stakeholders and holding your own in a commercial conversation about margin and trade-offs.
• Comfort operating as the sole data scientist in a domain — setting your own priorities and defending your own methodology.
• A degree in a quantitative field (Master's or higher preferred) or equivalent.

Nice-to-haves

• Direct experience building pricing or revenue-optimisation models — dynamic pricing, price elasticity, willingness-to-pay, discount optimisation, or risk-based pricing. You've seen a pricing model meet the real world and can talk about what broke.
• Multi-market or multi-currency pricing experience, especially in Southeast Asia.
• Feature store and streaming data experience.