Software Engineer, Data Platform

Job Description

The Data Platform Team provides platforms that support data analytic activities for all services offered on the LINE app.
We are currently looking for software engineers in the following three Parts: Ingestion Part (collects and aggregates data), Platform Dev Part (develops applications related to data analysis), and Hadoop Part (develops proprietary Hadoop clusters).

Our Mission

Ingestion Part

• Develop next-gen data processing and pipeline for the LINE messaging platform
• Design and build a distributed data processing system, such as Kafka, Elasticsearch, Hadoop, etc.
• Build a large-scale data pipeline for real-time data analysis
• Design and build an automated and reliable system in order to provide highly available services

Platform Dev Part

• Work closely with other teams (e.g. Ingestion Part and Hadoop Part) to develop tools for distributed systems in order to support the tasks of internal planners and data scientists
• Develop API servers to connect systems
• Revamp and operate the above tools and services 

Hadoop Part

• Develop a proprietary Hadoop distribution and its ecosystem in order to operate Vanilla Hadoop clusters consisting of a few thousand servers in a stable and effective manner
• Develop a security system to safely maintain confidential data
• Automate operation and build monitoring systems to further refine the platform, instead of using Hadoop out-of-box

About the Team

The Data Platform Team is responsible to collect data from LINE services, format data for analysis, and provide formatted data as a tool on its distributed system. Members of this team are expected to communicate and work closely with various departments, including service developers, Hadoop engineers, and data scientists. This team also communicates frequently with colleagues in both domestic and overseas offices. The team consists of members with different nationalities where communication is mainly in Japanese with occasional use of English. As with other departments, English is used in documentations and chat conversations. This team develops a wide range of tools for distributed systems (e.g. Spark and Presto) as well as more traditional tools for batch processing, admin screens, etc. Refer to the following for specific applications and peripheral systems developed by the Data Platform Team.

Announced at LINE Developer Day 2018 (November 2018)

• https://linedevday.linecorp.com/jp/2018/#A-5
• https://linedevday.linecorp.com/jp/2018/#B-6

Presented at LINE Developer Meetup in Tokyo #27 (January 2018)

• https://logmi.jp/tech/articles/286011
• https://logmi.jp/tech/articles/286477

Responsibilities

Ingestion Part

• Design, develop, and operate the data pipeline platform
• Design and develop a platform (e.g. Fluentd and Filebeat) to collect a large amount of data from servers and clients on the LINE platform
• Design and build large-scale ETL platforms (e.g. Kafka and Flink) to take collected logs and format, process, and save in storage, such as Hadoop and Elasticsearch
• Maintain the above platforms and provide user support
• Design and develop a system to provide the above log collection platforms and ETL platforms as internal SaaS

Platform Dev Part

• Work with planners to design, develop, and operate tools that fulfill internal requirements
• Provide user support
• Work with other teams to conduct troubleshooting when necessary

Hadoop Part

• Develop tools to operate and monitor clusters
• Develop LINE's proprietary Hadoop ecosystem
• Fix bugs on Hadoop and each software on the Hadoop ecosystem 

Current Product Phase/Exciting Challenges/Opportunities

The analytic platform offered by the Data Platform Team has completed the prototype phase and is currently being used by internal employees. The current focus is to develop new features, ensure scalability, enhance reliability, and automate operations in order to respond to various internal demands. Developing new features is not limited to simply adding features—this team, if needed, also takes up large-scale development to rebuild a feature from scratch. This means that you will have the opportunity to work with our large-scale data system from top to bottom.

Tools/Development Environments

• Streaming systems - Kafka/Flink
• Data collection tools - Fluentd/Filebeat/Logstash/Elasticsearch
• Hadoop ecosystems - HDFS/YARN/Hive/Presto/Spark/HBase
• Operating/monitoring tools - Kubernetes/Ansible/Grafana/Prometheus +
• Promgen/imon - internal monitoring tool
• BI - Tableau
• Development environments - IntelliJ/Eclipse/Github/Jenkins
• Programming languages - Java/Kotlin/Scala/Python

Qualifications

•  BS/MS degree or higher in Computer Science or Informatics (or equivalent work experience)
• Strong fundamental knowledge of computer science, such as data structure, algorithm design, and computational analysis
• At least 3 years of hands-on software development experience with Java
• Experience with concurrent/multi-threaded programming
• Experience with development and system operation in Linux/Unix environments
• Ability to set up machines using provisioning tools, such as Ansible
• Ability to set up monitoring

Preferred Qualifications

• Technical knowledge and competency to analyze, debug, and optimize large-data middleware, such as Kafka and Hadoop
• Ability to design, analyze and solve problems related to large-scale systems or distributed systems
• Experience in designing data pipeline platform
• Understand the semantics of distributed data pipeline (e.g. 'at-least-once' and 'exactly-once') and have the technical capability to build systems
• Proficient at data analytic engines, including Elasticsearch, Hadoop, Spark, and Presto
• Proficient in data collecting tools, including Fluentd, Embulk, and Filebeat
• Experience developing or operating frameworks and platforms
• Experience building a system leveraging container-related technologies, such as Kubernetes
• Knowledge of relational database engines
• Experience with provisioning mission critical, 24/7 systems
• Experience with troubleshooting/tuning JVM GC
• Experience with Maven, Gradle, and Git 

Ideal Candidate

• Technical knowledge and competency to analyze, debug, and optimize large-data middleware, such as Kafka and Hadoop
• Ability to design, analyze and solve problems related to large-scale systems or distributed systems
• Experience in designing data pipeline platform
• Understand the semantics of distributed data pipeline (e.g. 'at-least-once' and 'exactly-once') and have the technical capability to build systems
• Proficient at data analytic engines, including Elasticsearch, Hadoop, Spark, and Presto
• Proficient in data collecting tools, including Fluentd, Embulk, and Filebeat
• Experience developing or operating frameworks and platforms
• Experience building a system leveraging container-related technologies, such as Kubernetes
• Knowledge of relational database engines
• Experience with provisioning mission critical, 24/7 systems
• Experience with troubleshooting/tuning JVM GC
• Experience with Maven, Gradle, and Git 

Compensations

Annual salary system (To be determined based on skills, experiences and abilities after discussions)

• Annual compensation will be divided into 12 months and paid on a monthly basis.
• Separate incentives available (*1)
• Compensation revision: twice a year
• Allowances: commuting allowance, LINE Pay Card Benefit Plan (*2)
  (*1) In addition to your annual compensation, you may receive incentives (twice a year) depending on the company's and individual performance and evaluation on your performance. (Incentives are not guaranteed to be provided. An incentive payment will only be paid if you remain employed as of the payment date.
  (*2) This is an allowance separate from the salary meant for employees to use for their health, personal development, support for raising the next generation, and more.