This role focuses on building experimentation frameworks and personalization systems that optimize the property discovery experience and drive measurable business outcomes across the real estate platform. You will design and execute experiments, develop personalization models, and create data products that adapt to individual user preferences and behaviors in real-time.

You will lead experimentation initiatives from hypothesis generation through analysis, while building sophisticated personalization engines that continuously learn from user interactions. The role requires expertise in causal inference, experimentation design, and personalization at scale—translating complex user behavior into tailored experiences that increase engagement and conversion.

Experimentation:

• Design and execute A/B tests, multivariate tests, and sequential experiments across search, recommendations, and user journeys
• Build experimentation frameworks and statistical methodologies to measure impact on engagement, conversion, and revenue metrics
• Develop causal inference models to understand treatment effects and optimize allocation strategies
• Implement multi-armed bandit algorithms and contextual bandits for dynamic experimentation
• Define success metrics, sample size requirements, and guardrail metrics for experiments
• Conduct post-experiment analysis including heterogeneous treatment effects and long-term impact assessment

Personalization:

• Design and deploy real-time personalization systems for property search, homepage content, alerts, and recommendations
• Build user segmentation models and dynamic preference profiles based on behavioral signals, explicit feedback, and contextual data
• Develop multi-objective optimization models that balance relevance, diversity, and business goals
• Create adaptive ranking models that learn from implicit feedback (clicks, time-on-page, saves) and explicit signals
• Implement cold-start strategies for new users and properties using contextual information and market data
• Build propensity models for user intent (likelihood to contact, convert, churn) to enable proactive personalization

Key Responsibilities:

• Design, build, and evaluate recommendation systems for property search, property alerts, similar listings, and personalized homepages
• Develop personalization models using user behavior, preferences, location, and property characteristics
• Use machine learning techniques such as collaborative filtering, content-based filtering, ranking models, and clustering
• Build predictive models for user intent (e.g., likelihood to contact agent, save a property, schedule a viewing)
• Work with large datasets: user activity, property listings, interactions, and market data, ensuring high data quality
• Perform feature engineering to represent property attributes, neighborhood data, and user signals
• Deploy models into production through APIs or batch pipelines and support A/B testing of model performance
• Build dashboards and reports to monitor model impact and user engagement
• Communicate insights and model results to product managers, engineers, and business stakeholders
• Stay updated on advances in recommendation systems and machine learning best practices

Skills:

• Strong proficiency in Python (Pandas, NumPy, Scikit-learn, PyTorch/TensorFlow) and SQL
• Experience building recommendation and personalization systems
• Strong knowledge of machine learning, predictive modeling, and statistics
• Experience with feature engineering and model evaluation (precision/recall, NDCG, CTR, conversion rate, etc.)
• Familiarity with experimentation and A/B testing frameworks
• Experience working with large-scale, structured and unstructured data
• Data visualization and communication skills are a plus

Qualifications:

• Degree in Computer Science, Statistics, Mathematics, Engineering, or related field
• 5+ years of experience in data science with a focus on machine learning or recommendation systems
• Experience in real estate, marketplaces, e-commerce, or classifieds platforms is a strong plus
• Experience with cloud platforms (AWS is preferred)