Data Scientist

Papaya is a leading skill-based mobile gaming company, bringing together fun, competition, and real rewards for millions of players worldwide.

Driven by our vision to create one of the world’s most exciting player communities, we develop games powered by a large-scale B2C platform that supports tens of millions of daily tournaments and connects players around the world through social competitions.

Located in the heart of Tel Aviv, we offer a fast-moving environment, culture of innovation, professional growth, and the opportunity to make a true impact.

We run real-money skill-based games where many product decisions are prediction problems: who will churn, who will deposit again, what a fair match looks like, and whether a marketing cohort will pay back. We’re hiring a Growth Data Scientist to replace heuristics with production models that directly influence what players see and do.

This is not a research role—you’ll own problems end to end, from problem definition and labeling to data, modeling, backend integration, and measuring impact.

Responsibilities:

• Cohort LTV & Churn Prediction. A per-player risk score for depositors, labelled against repeat-deposit KPIs—wired to real interventions, and cohort-level LTV:CAC maturation modelling to tell marketing in-month whether a cohort will pay back within a year.
• CRM & Lifecycle Optimization. Building models to trigger personalized Push and Email notifications at the exact right moment, matching the right communication journeys to the player's specific lifecycle stage and behavior.
• Deposit-after-N-days-without-deposit probability. A conditional/hazard model for the likelihood a lapsed depositor deposits again given N days of inactivity — feeding RV model targeting and treatment timing.
• Lobby suggestion. Which tournaments and which entry-fee tiers to surface to which player. Today the lobby is driven by hand-built segmentation rules; we want a learned ranking model that trades off engagement, monetization, and matchmaking liquidity.
• Store suggestion. Personalized offer and pack ranking in the store, replacing static store-segmentation configs — picking the right offer, at the right price point, at the right moment.

Requirements:

• 4+ years of applied data science in growth, monetization, retention or lifecycle — with models you personally got into production and can point to a business decision they changed.
• Prior domain experience is required, not a bonus: mobile gaming, real-money gaming/iGaming, or a consumer app with a live in-app economy. You should already think in LTV, ARPDAU, D-n retention, offer elasticity, entry fees and payback windows without being taught the vocabulary.
• Hands-on propensity, churn, uplift, survival/hazard and LTV modelling. Recommendation or ranking experience is a strong plus.
• Python at production quality (LightGBM/XGBoost; PyTorch or TensorFlow for embedding-based work) and strong warehouse SQL (Snowflake, BigQuery).
• Real experimentation depth — designed tests, sized them, and read them honestly. Including at least one case where a clean A/B wasn't available and you still produced a defensible impact claim.