Job Description:

Business Overview 

Global Ad Technology Supervisory Department(GATD) manages the whole of the advertising systems that power Rakuten. We are a cross-functional and data-driven organization working together in a diverse team spread across five countries: Japan, Singapore, India, China and UK. 

GATD Website:

Department Overview 

Our vision is to empower our internal and external stakeholders by providing data driven marketing solutions by developing the Ad delivery, Audience Management, Tracking, Reporting and Data Sync Platforms. Joining us, you will be able to challenge yourself in a wide range of technologies in the frontline of heavy traffic large scale ad applications and platforms, including high performance system development and big data processing systems. Also, you will be able to work about latest digital advertising and Ad platform because we also closely work with outside Ad platformer as well.  

In the platform team, we have positions who have comprehensive knowledge of the platforms and provide total solution for business requirement by combining of multiple platform components. 

Our team develops and operates an ad serving platform that is widely used in Rakuten's overall services, including Rakuten Ichiba. The system consists of servers that respond to high traffic, such as tens of thousands of QPS, with low latency of less than 100ms, a backend system that processes huge log data of tens of billions of data per month, and campaign management UI. 

Position:

Why We Hire 

With the practical adoption of AI, decision-making in ad serving is becoming faster and more sophisticated, and competition among platforms is intensifying. In this context, our ad serving platform (RUNA) is in a phase of differentiating itself by strengthening the in-house development of core algorithms centered on price optimization, ad performance maximization, and creative optimization. In addition, RUNA's adoption within the Rakuten Group continues to expand, and as the scale of delivery and the diversity of use cases grow, we need to enhance model generalization, experiment velocity (A/B testing), and operational reliability. To address these external changes and rising internal demand, we are seeking data scientists who can drive initiatives end to end—from hypothesis formulation and modeling to online deployment, evaluation, and continuous improvement. 

Position Details 

- Problem definition for ad delivery optimization and the design of KPIs/metrics, along with observability (dashboards and alerts). 

- Data preprocessing, feature engineering, modeling, and offline evaluation (including simulation) for price optimization, CTR/CVR and value prediction, and creative optimization. 

- Online deployment and experiment design: A/B tests/bandits, calibration, rollout strategy planning, effectiveness evaluation, and continuous improvement. 

- Production operations and MLOps: building training/inference pipelines; monitoring and drift detection; model updates and automated rollback; quality and privacy governance. 

- Collaboration with internal stakeholders (product, engineering, sales/operations) on requirements definition, decision support, documentation, and knowledge sharing. 

Work Environment 

Development team

About 20 people (SRE, Backend/Frontend, Data, SDK, QA) 

Development Environment 

- Google Cloud Platform(GKE, Dataflow, Cloud Pub/Sub, BigQuery, Cloud SQL) 

- Aerospike 

- Terraform 

- Ansible 

- Python 

- Go

Related systems, advertising products

- Ad Platform / DMP / Data feed / Tracking and reporting / Pixel tag 

- Listing ads, Display ads, Video ads, and Google shopping ads 

Mandatory Qualifications: 

- 3+ years of hands-on experience in data analysis and machine learning using Python and SQL; proficiency with Pandas/NumPy/scikit-learn and data visualization tools 

- Experience processing large-scale data (e.g., Spark, BigQuery) and building reproducible analytics environments (using notebooks and repositories together, code reviews, testing) 

- Solid foundation in statistics, probability, causal inference, and experimental design (hypothesis testing, sampling, A/B test design and analysis, effect size estimation) 

- Experience developing machine learning models (regression, classification; tree-based methods/GBDT/logistic regression/simple neural networks) with rigorous evaluation design (offline/online metrics, calibration) 

- Foundational knowledge of online/digital advertising (CTR/CVR prediction, bidding and budget pacing, ad serving control, KPI optimization) or practical optimization experience in adjacent domains 

- Feature engineering oriented toward price and serving optimization use cases; data quality management, leakage prevention, and handling delayed/late-arriving labels 

- Production experience (model deployment, scheduling/pipeline operations, monitoring, implementing and operating drift detection) 

- Communication skills for collaborating with stakeholders on requirements definition, analysis design, and decision support (including explaining and visualizing deliverables) 

- Degree in computer science, information engineering, statistics, applied mathematics, or a related field; or equivalent practical experience 

- Fluent English, or a TOEIC score of 800+ (or an equivalent qualification/skill) 

Desired Qualifications: 

- Knowledge of ad auctions and mechanism design (first-price/second-price, bid shading, reserve price optimization, frequency control/capping) 

- Hands-on experience applying price optimization methods (Bayesian optimization, multi-armed bandits, contextual bandits, reinforcement learning, etc.) 

- Experience with creative optimization (multivariate testing, bandit optimization, image/text feature extraction, use of generative AI) 

- Experience building serving-control models and simulations (e.g., budgeting, pacing, inventory forecasting) 

- Experience implementing and operating streaming/real-time processing platforms (Kafka/Flink/Beam) and online (real-time) inference 

- MLOps practices (feature stores, model registries, CI/CD, Docker/Kubernetes, monitoring/alerting, data drift/concept drift detection) 

- Experience with cloud environments (GCP: BigQuery/Vertex AI/Dataflow, or equivalent services on AWS/Azure) 

- Professional experience leveraging AI in real-world applications 

- Fluent Japanese speaker or have N2 Level of the Japanese Language Proficiency Test, or have equivalent qualifications or business level skills 

#engineer #applicationsengineer #globaladdiv #RakutenIchiba