Lead Data Scientist, Insurance

Overview

Traveloka's insurance division operates at the crossroads of travel demand and risk management, aiming to offer timely, precise insurance products alongside travel purchases such as flights, hotels, and activities. The pricing mechanism is a crucial lever influencing both attachment rates and profit margins.

Role and Responsibilities

• Take full ownership of the insurance pricing domain by collaborating with product managers, data analysts, and engineers to translate business challenges into dynamic pricing models that facilitate strategic decision-making.
• Design, train, and regularly update pricing models by curating specific training datasets using SQL, performing exploratory data analysis, and selecting relevant features.
• Perform extensive validation including simulations and backtesting to benchmark model performance, assess revenue impact, and identify potential issues prior to deployment.
• Deploy models into production and maintain their operational health with assistance from Data Engineering, Data Analytics, and the ML Platform teams, enabling experimentation across various pricing strategies such as revenue maximization and seasonal discounts.
• Communicate methodologies, trade-offs, and findings clearly to both technical teams and non-technical stakeholders, translating results into actionable leadership recommendations.
• Lead and define the future direction, methodologies, and standards of pricing science, engaging stakeholders throughout the process.

Required Qualifications

• Over six years of professional experience in data science or applied machine learning, managing end-to-end ML projects.
• Expertise in deploying models that serve live production traffic with real-time inference considerations like feature freshness and fallback mechanisms.
• Proficiency in SQL and Python, with strong knowledge of modeling techniques including gradient boosting, regression, and causal inference methods, combined with sound judgment in model complexity selection.
• Experience in experimental design and causal inference techniques such as A/B testing, holdout groups, backtesting, capable of distinguishing true model improvements from seasonal effects.
• Familiarity with software engineering principles including version control, CI/CD pipelines, containerization, basic cloud technologies, and monitoring to work independently with platform support.
• Ability to collaborate and negotiate directly with business stakeholders on margin impacts and strategic trade-offs.
• Comfort in functioning as the sole data scientist for a domain, independently setting priorities and defending methodological choices.
• Bachelor’s degree in a quantitative discipline is required; a Master’s degree or higher is preferred.

Preferred Experience

• Hands-on experience building pricing or revenue optimization models encompassing dynamic pricing, price elasticity, willingness-to-pay analysis, discount strategies, or risk-based pricing, including practical insights into real-world challenges.
• Exposure to multi-market or multi-currency pricing environments, particularly in Southeast Asian regions.
• Knowledge of feature stores and streaming data architectures.