Chin-Hung (Wilson) Liu
I'm leading, architect a large-scale data pipeline from Paktor x 17LIVE (GCP/AWS/python/scala), cooperating with Singapore data science/machine learning team/TW HQ data team, and Hadoop ecosystem related experiences such as HDFS/Hbase/Kafka in JSpectrum (Hong Kong, Sydney).
● Having 12+ years of experience in designing, developing various Java/Scala/Python based applications that support day to day operations.
● 6+ years of working on multiple issues as a data team member or driving analysis and elaboration, design, development of data-pipeline applications and builds automation tools.
● Extensive knowledge of Spark and Hadoop ecosystem, such as Hadoop, HDFS, HBase, etc.
● Strong knowledge in developing/designing AWS/GCP/Azure Big Data services.
● Extensive skill towards developing generic distributed systems, streaming processing, deployed machine learning pipeline and continuous development of ML models.
Senior Data Engineer at Paktor x 17LIVE| AWS x GCP x Azure Big Data Specialist
Singapore / Hong Kong / Taiwan
https://www.linkedin.com/in/chin-hung-wilson-liu-29392957
Nanxing Rd., Xizhi Dist., New Taipei City, Taiwan (R.O.C.)
Experience
Senior Data Engineer / Handshakes by DC Frontiers / Taiwan, Taipei.
Oct. 2020 - Present.
- Identify, digging bottlenecks, and problem-solving especially optimizing the performance of SQL Server, NoSQL (Azure Cosmos), resource units, and message queues, reducing/saving almost 50-75% of resources.
- Identify, solving the problems between machine learning/backend/frontend/DDP side and giving the advance logical/physical design of a system.
- Displayed technical expertise in optimizing the databases and improving the data pipeline to achieve the objective.
- Bring in industry standards to data management to delivery of data at the end objective.
- Build, recruiting the new data engineering staff for the next-generation, enterprise data pipeline.
- Storage, Azure Cosmos DB/Gremlin/SQL Server/MYSQL/Redis
- Storage (Bucket), Azure Blob/AWS S3
- Streaming/Batch/transform, Spark/Scala (90% codebase coverage)
- Message, Azure service bus, queue storage
- Search, Elasticsearch
- Algorithm, graph/concordance
Senior Data Engineer / M17 Entertainment Group / Taiwan, Taipei.
Feb. 2020 - Jul. 2020
Description and Responsibilities : The big challenge of 17 Media data teams is facing fast-growing data volume (processing 5-10x TB level daily), complex cooperation with stakeholders, the cost optimization of pipeline and refactor big latency systems .etc. As a senior data member, I’m making a data dictionary and trying to explain/design how the whole pipeline working with each components, especially how to solve those bottlenecks. More responsibilities / details as below.
- Leading, architect a large-scale data pipeline for supporting scientists and shareholders. - Optimize, ensure quality and play a tough role in data lake projects / data pipes. infrastructure.
- Define, designing stage, dimension, production and fact tables for data warehouse (BigQuery).
- Coordinate with client / QA / backend team for QC lists / MongoDB change stream workers.
- Architect workflows with those components, Dataflow, Cloud Functions and GCS.
- Recruiting (Jr./Sr.) data engineering members, setting goals and sprint management.
Tech Stacks :
- Storage, GCS/BigQuery/Firebase/MongoDB/MYSQL
- Realtime process and Message system, DataFlow (Apache Beam) / BigQuery Streaming / MongoDB Change Stream / Fluentd / Firebase / Pub/Sub
- ETL/ELT workflow, Digdag / Embulk
- Data warehouse, Visualization, BigQuery / Superset / Chartio / Data Studio
- Continuous deployment, docker, CricleCI
Reports to : Data Head
Data Engineer / Paktor Pte. Ltd. / Singapore
Sep. 2015 - Dec. 2019.
Description and Responsibilities : This is another 0 to 1 story. As an early data member, we need to figure out the data driven policy, strategies, engineering requirements from the company. In Paktor, data / backend sides are 100% on AWS, therefore the whole data ingestion, automation and data warehouse etc. are relying on those components. We are processing 50-100x GB realtime / batch jobs and the other data sources (RDBMS, APIs) for ETL/ELT on S3, Redshift, the data platform helps our marketing / HQ scientists team getting data into insights and making good decisions. More responsibilities / details as below.
- Supports Big Data and batch, real-time analytical solutions leveraging transformational technologies.
- Optimize data pipeline on AWS using Kinesis-Firehose/Lambda/Kinesis Analytics/Data Pipeline, and optimize, resizing Redshift clusters and related scripts.
- Translates complex analytics requirements into detailed architecture, design, and high performing software such as machine-learning, CI/CD of recommendation pipeline.
- Collaborate with client / backend side developers to formulate innovative solutions to experiment and implement related algorithms.
Tech Stacks :
- Storage, S3/Redshift/Aurora - Realtime process and Message system, Kinesis Firehose / SNS
- Data warehouse, Visualization, Redshift / Klipfolio / Metabase
- ETL/ELT workflow, Lambda / SNS / Batch / Python - Recommendation, ML, DynamoDB / EMR / Spark / Sagemaker
- Metadata management, Athena (presto) / Glue / Redshift Spectrum
- Continuous deployment, Elasticbeanstalk / Cloudformation
- Operations, PagerDuty / Zapier / Cloud Watch
Reports to : CTO, Data Head
System Analyst (Data Backend Engineer) / JSpectrum Software Limited / Hong Kong
Jan. 2014 - Aug 2015.
Description and Responsibilities : JSPectrum is a leading passive location-based service company in Hong Kong which holds many interesting products such as NetProbe, NetWhere, NetAd etc. In Optus (The main project in Sydney), the main responsibility of system analyst is designing / implementing data ingestion (real-time processing) / load and management data with major components of the Hadoop ecosystem. We meet the challenge to process 15,000 TPS, 60,000 inserts per second and 300 GB daily storages, therefore we are trying to optimize those components with Kafka consumers, HDFS storages and re-designing keys / columns of HBase to fulfill the requirement and deployed NetAd, whole in-house solutions on Optus. More responsibilities / details as below.
- Design, implement and optimize Hadoop ecosystems, MLP, real-time processing on Optus in house servers with our main product NetAd, NetWhere. We are focusing on HBase schema, HDFS, balancing Kafka consumers and more issues on data ingestion.
- Collaborate with shareholders and LBS team members for further requirements with HeapMap.
Tech Stacks :
- Storage, HDFS / HBase
- Realtime process and Message system, Kafka streaming, Log systems
- Data warehouse, Visualization, HBase / NetWhere (Dashboard)
- Hadoop ecosystem, Hadoop / HDFS / Zookeeper / Spark / Hive
- ETL/ELT workflow, Spark / Hive / Scala / Java
Reports to : CTO
Senior Software Engineer / Toro Development Ltd. / Taiwan, Taipei.
Oct. 2012 - Dec. 2013.
Description and Responsibilities : TORO is a technology business that provides a mobile platform and its associated systems, services and rules to help Brands (with initial focus on Sports Teams, Smart Cities and Streaming apps) become super-apps to generate additional revenue with minimum effort. Responsibilities as below.
- Design, implement and test back-office modules for NFC wallet platform, Trusted Service Managers (TSM) and distributed NFC services to end users / stakeholders.
- Implement RESTful services and deliver endpoints for wallet managers and collaborating with frontend, backend teams for further business requirements.
Tech Stacks: MYSQL / Spring / Hibernate / XML / Apache Camel / Java / POJO .etc.
Reports to : Head of Server Solutions
Software Engineer / Digital River / Taiwan, Taipei.
Oct. 2011 - Sep. 2012.
Description and Responsibilities : Digital river proactive partners, providing API-based Payments & Risk, Order Management and Commerce services to leading enterprise brands. The big challenge to DR is integrating with the current module and working well with a huge code base (over 2+ millions lines), the strict process including analysis requirements, design, implement, test and code review. More responsibilities as below.
- Design, implement custom bundle project, bundle customized by shoppers to pick products of groups and get special discounts, the main stakeholders /users from Logitech, Microsoft.
- Analysis, collect business requirements, identify use cases and collaborate with business analysts and deliver related diagrams, documents.
Tech Stacks: Oracle / Tomcat / Spring / Struts / JDO / XML / JUnit / Java / J2EE .etc.
Reports to : Technical Development Manager
Technical Supervisor / Stark Technology Inc. / Taiwan, Taipei.
Oct. 2008 - Sep. 2011.
Description and Responsibilities : Stark Technology (STI) is the largest domestic system integrator in Taiwan. We plan and deliver complete ICT solutions for a wide spectrum of industries through representing and reselling the world's leading products. This is made possible by using the most advanced technology, and providing the best professional services. More responsibilities / projects as below.
- Lead, coach JR. programmers for the development process of enterprise modules, and design Fatwire CMS components as Template/Page/Cache .etc.
- Design, analyze DMDB systems, and implement functions to meet the requirements of queries / storage. Optimize performance for online servers and GC tuning.
Tech Stacks : Oracle / Sybase / Tomcat / Weblogic / Spring / Struts / Hibernate / Fatwire / Java / J2EE .etc.
Reports to : Technical Manager
Relevant Skills and Qualifications
Big Data Tech Stacks
- AWS Services, EC2/S3/Lambda/EMR/CloudWatch/SNS/Elastic Beanstalk
- AWS Big Data Solutions, Kinesis/Firehose/Athena/Redshift/Dynamodb
- GCP Big Data Solutions, BigQuery/PubSub/Dataflow/Cloud Functions
- Hadoop ecosystem, Hadoop/HDFS/Zookeeper/Hbase/Hive
- Spark Streaming/Apache Kafka
- CI/CD: Jenkins/Cloud Formation/Maven/Git/Gradle
- Solid, well-designed real-time streaming/batch processing, ETL systems.
- Monitors and conducts data-pipeline / machine learning pipeline development requests through lifecycle management and ensures that the technical solution meets.
- Diagnosing and troubleshooting Redshift and specific clusters management.
- Development of micro-services and endpoints based on enterprise integration patterns. Knowledge over garbage collection (JVM) tuning technologies for various servers.
- Developed multi-threading processing consuming work and managed transactions.
Certifications and Training
- Sun Certified Web Component Developer Java 2 Platform, Enterprise Edition.
- Sun Certified Programmer for the Java 2 Platform.
- Red Hat Enterprise Directory Services and Authentication Attended.
- Project Management Professional (PMP)® Attended.
- AWS Certified Solutions Architect Attended.
- Big Data on AWS Attended.
- Azure Data Engineer AssociateAttended.
Education
National Taiwan University, 2010 – 2011
EMBA Programs, Business Administration, Accounting, Finance and International Business.
Chinese Culture University Master of Information Management, 2002 – 2005
Computer Science, Data Mining, Expert Systems and Knowledge Base as major concentration.
Chinese Culture University, Bachelor Degree of Science in Journalism, 1998 - 2002
