G Talent Top
Jobs
Data Scientist / Hybrid Work / Major Internet Service

Data Scientist / Hybrid Work / Major Internet Service

Tokyo

Fulltime

Remote

Large Scale Company

Listed Company

Own Products/Services

Global Business

Many Foreign Employees

Industry: Major Internet-related Service Provider

IT Skills: Java, Pandas, Python, Numpy

Working hours: Fixed Time System

Salary: 6 Million yen〜10 Million yen

Job Description: The Group R&D Division is a department that provides technical support, development, and analysis for startups and cross-group projects that the company is focusing on in its business areas to support business success. Within this department is the AI R&D Office, which provides support related to data analysis and AI. In addition, they are constantly researching and developing the latest technologies, and are quick to introduce them into businesses to achieve results. You will first participate in one of the following. You will then either further develop your expertise or join other projects to develop your desired career depending on your performance. ◆Adtech Projects In RTB (Real Time Bidding), one of the main mechanisms of Internet advertising, we mainly design and develop machine learning models for DSP (Demand-Side Platform), which optimizes cost-effectiveness for advertisers, and measure their effectiveness. ◆App projects Improve KPI of the service by measuring the effectiveness of new features and measures of the application that facilitates free WiFi connection using causal inference techniques and setting up a system for data-driven management decision making. ◆Other Projects Support by applying data analysis and machine learning techniques with respect to crypto asset trading, fraud detection, etc. 【R&D work】・While performing project work, all members spend a certain amount of time researching cutting-edge machine learning methods and new machine learning applications. ・In addition, members appointed on a quarterly basis conduct focused research and development. 【Utilized Technology】 ◆Analysis Methods Machine learning: Transformer systems (large-scale language models, etc.), graph neural networks (GNN), multilayer perceptron (MLP), ensemble learning/gradient boosting (Gradient Boost Tree + LR, Random Forest, ExtraTree , Ada Boost XGBoost, LightGBM), PCA, FP-Growth, Word2Vec, Doc2Vec, collaborative filtering, Bayesian estimation, HMM models (hidden Markov models) Statistical Analysis: t-test, Chi-square test, F-test, binomial test, Kolmogorov-Smirnov test, Shapiro-Wilk test, sampling (MCMC, bootstrap method, etc.), analysis of variance, causal inference (difference in differences method, etc.) 【Development Technology/Environment】 Programming/Frameworks ・Python, PyData (numpy, scipy, pandas, etc.), Streamlit ・PyTorch, TensorFlow, LangChain, Spark (PySpark) Cloud/On-premise (middleware) ・Google Cloud (GCS, BigQuery, VertexAI, Dataflow, etc.) ・AWS (S3, Athena, EMR/Serverless, StepFunction, SageMaker, Bedrock, etc.) ・MySQL, MariaDB, Percona Server, PostgreSQL, Galera Cluster, Oracle, Hive, Hadoop/HDFS ・ConoHa (GPU server) Large-scale language model (LLM) related OpenAI API, Llama3, LangChain, HuggingFace ◆Development tools ・Atlassian (Jira, Confluence), Trello ・VS Code, PyCharm, Jupyter ・GitHub (Copilot) ・Tableau, Looker Studio, metabase ・ChatGPT, Gemini, Claude ◆Development Methodology Agile development (scrum-based) 【Attractiveness of the Position】 You will have the opportunity to launch projects from scratch to solve various issues through data analysis and AI technologies for the company group's diverse services. You will be able to directly handle a wide range of data, including the world's largest financial data and several hundred terabytes of ad tech data, and learn big data analysis technologies (BigQuery, PySpark, etc.). You will be able to analyze various types of data such as time-series data, user behavior data, and articles. You will constantly learn about the business domain of the projects you are in charge of, as well as cutting-edge machine learning, deep learning, and statistical methods. You will acquire three important skills (business problem solving, data science, and engineering) and can greatly develop one or more of your strengths. You will have many opportunities to interact with data scientists from different departments who are in charge of different projects through study sessions and other opportunities to enhance your data science skills. You will work together with members of the Next Generation System Laboratory, an elite group of engineers under the Group's Research and Development Division, and you will be able to develop your engineering skills. As a department under the direct control of the group CTO, the selection of technologies is left to the field, allowing you to verify and introduce cutting-edge technologies on your own. Since all services are in-house, you will be able to work with the business unit to set issues and come up with solutions on your own, and carry out improvement cycles based on data science. Depending on the individual's performance and motivation, you will be able to take on the challenge of freely conducting research and development work on the team's key themes. The team is diverse, consisting of PhDs who have worked in academic fields, engineers, and others.

Required Skills: 【Required】 Meet all of the following ■Experience using machine learning/deep learning in actual products or research to produce results ■Knowledge and skills in information systems, with a full range of skills required by data scientists (statistical analysis, machine learning, etc.), or one specialized strength ■Able to write Python 【Preferred】 ■Ph.D. degree. ■Experience as a consultant or in the business sector 【Ideal Applicants】 ■Likes to consider all approaches to solving business problems and finding the best way to do it, rather than just using a method. ■Interested in and passionate about anything and everything, and able to take on new technologies and new tasks ■Able to work as a team to achieve great results that cannot be done alone ■Critical thinking, not just relying on data and results

Required Language Skills