Senior Data Scientist

About Scrambly

Scrambly is a rapidly expanding adtech startup recognized in the AppsFlyer Performance Index and Singular ROI Index within just 3.5 years. The company is growing over 250% year-over-year in revenue, team size, product development, and technology, operating profitably and without external funding. Scrambly is developing an innovative, reward-enhanced alternative to traditional app stores, linking millions of users to top apps and enabling advertisers to achieve exceptional ROI through a loyalty-based ecosystem.

Role Overview

We seek a Senior Data Scientist to lead the development and maintenance of models that drive Scrambly's financial decisions, such as user valuation, acquisition bidding, and fraud prevention. Your primary focus will be predicting individual and group user value by interpreting early behavioral data and continuously refining these forecasts as market conditions and products evolve. This role encompasses interconnected challenges spanning marketing, monetization, and risk management, translating predictive insights into actionable business strategies. Collaboration with engineering teams for data infrastructure and integration is essential.

Key Responsibilities

• Develop and enhance models forecasting user and cohort lifetime value across multiple revenue channels, focusing on commercial impact and model robustness amid changing products.
• Construct bidding and pricing models that guide user acquisition spending, including mechanisms to maintain budget limits during the learning phase.
• Create segmentation and scoring systems that categorize users, partners, and inventory according to their actual contribution value.
• Refine fraud detection models to improve precision, balancing fraud prevention with minimizing false penalties on legitimate users.
• Establish and uphold rigorous model evaluation standards including out-of-sample testing, backtesting, and metric selection aligned with business priorities, with ongoing monitoring and alerting on model performance decline.
• Collaborate closely with the Head of Data on project prioritization, engineering teams on pipelines and model delivery, and commercial groups to build trust in model outputs for decision-making.

Candidate Profile and Requirements

• Minimum four years of experience as a data scientist creating production-grade models that inform real business decisions, emphasizing ownership and impact over tenure alone.
• Expertise in lifetime value (LTV) or return on ad spend (ROAS) prediction, particularly handling immature cohorts, extended forecasting horizons, sparse segments, and model degradation due to product shifts.
• Experience in adtech or mobile performance marketing domains, with deep understanding of attribution, mobile measurement partner (MMP) data, cohort economics, and user acquisition spend dynamics. Backgrounds in gaming, gambling, or user incentive and loyalty programs are also relevant. Familiarity with rewarded, offerwall, or incentivized app ecosystems is a plus.
• Proficient in Python and SQL with experience deploying models in cloud-based data warehouses, preferably Google BigQuery.
• Strong commitment to rigorous model evaluation, selecting metrics that directly align with business goals, and careful validation including out-of-sample and point-in-time assessments. Prepared to reject models that do not meet quality thresholds.
• Proven track record of successfully deploying models into production environments and understanding the post-deployment lifecycle and maintenance.
• Sound commercial insight and effective communication skills to explain model behavior and outcomes to stakeholders who rely on these insights for budgeting and spending decisions, with the ability to challenge unsupported requests.
• Self-motivated and autonomous work style suited for a small team with minimal formal processes.
• Proficient English communication skills (B2 level or higher), both written and verbal, for documentation and interaction with a global team.

Preferred Qualifications

• Experience with bidding, pricing, and budget allocation strategies including control systems, bandit algorithms, or other approaches for managing spend under uncertainty.
• Expertise in fraud detection, anomaly identification, or trust and safety modeling, particularly with incomplete or weak labeling.
• Familiarity with experimental design and causal inference techniques such as A/B testing, power analysis, and incrementality measurement.
• Knowledge of advanced statistical methods including survival analysis, hierarchical modeling, or Bayesian techniques applied to retention, monetization, or segment-sparse problems.
• Experience with model development close to data warehouses using tools like BQML, Dataform, or dbt.
• Hands-on capability deploying and maintaining models including building pipelines for serving, retraining, and monitoring data/model drift.

What We Provide

• Direct responsibility for the models governing user acquisition spend, reward offers, and fraud management, offering rapid and visible impact.
• Exposure to complex challenges including long-term forecasting, adaptive modeling, pricing amidst uncertainty, and fraudulent behavior detection with imperfect data.
• Close partnership with the Head of Data, significant influence on the growing data function, and professional growth opportunities as the team scales.
• Flexible seniority and compensation calibrated to your expertise rather than fixed titles.
• Join a profit-positive, bootstrapped company experiencing rapid growth where data investments are driven by measurable outcomes.

Additional Information

By applying, you consent to the collection, processing, and storage of your personal information solely for recruitment and evaluation purposes by the employer.