Benjamin Kim

520 S. Burnside Ave. #7D, Los Angeles, CA 90036 (818) 635-2900

 

Education

 

University of California, Los Angeles, Los Angeles, CA              1994

Bachelor of Science, Microbiology and Molecular Genetics

 

Certifications

Cloudera Admin

 

Technical Skills

 

Data Systems:              Cloudera CDH [Flume-NG, HBase, HDFS, Hive, Hue, Impala, MapReduce, YARN, Oozie, Sqoop, Zookeeper, Spark], Amazon AWS [S3, EC2, Elastic MapReduce], Cloud Compute [Storage, BigQuery].

Applications:              Pentaho, Talend, Puppet, and Ganglia.

OS:                            Mac OS X, RedHat Enterprise Linux, CentOS Linux, and Ubuntu Linux.

Scripting:              Bash, Hive QL, PL/SQL, and Scala.

 

Professional Summary

 

Big Data Manager with 13+ years of Data Platform experience including:

 

·         Advocating close to real-time processes using the tools available within CDH, such as Flume-NG, HBase, HBase Coprocessors, Impala, and the upcoming Spark integration to automate online feedback mechanisms.

·         Blueprinting, documenting, and supervising the implementation of the big data infrastructure using open source technologies utilizing Cloudera’s ecosystem of software included in CDH4 such as Cloudera Manager, Hadoop, Hue, Beeswax, Oozie, Pig, Hive, Impala, and HBase to process hundreds of Terabytes of information.

·         Experience with project management, requirements documentation, client-relations, ETL workflow tools, and software development lifecycle methodologies including waterfall, agile, scrum, and kanban.

·         Strong understanding of Amazon AWS as an IaaS for S3 storage, Hadoop/Hive for analytics on EMR, and PostgreSQL for the dashboard in Microstrategy.

·         Responsible for overseeing the batch data workflow processes created using Spring Batch, Hive QL against Hive and Impala, Java against HBase, and direct Shell commands; coordinate efforts with Operations and Infrastructure to verify that the cluster nodes are configured and performing in accordance to user expectations.

·         Troubleshooting any issues in the Hadoop ecosystem by triaging any and all issue tickets; consulted and assisted in researching both configuration and implementation related issues, which included code reviews of Hive QL and Java; and forwarding all remaining blockers to the experts both within the company, online, or to Cloudera.

·         Led and facilitated a proof of concept ETL solution to populate HDFS using the open source data integration tool from Talend.

 

Experience

 

Adconion, Santa Monica, CA              09/2012 – Present

Big Data Manager

 

·         Pushing the company into transitioning to close to real-time processes using the tools available within CDH, such as Flume-NG, HBase, HBase Coprocessors, Impala, and the upcoming Spark integration; this system can unify and simplify all data pipelines and enable the company to automate a feedback mechanism, which will configure campaign parameters continuously, using online analytics.

·         Blueprinted, documented, and supervised the implementation of the big data infrastructure using open source technologies using Cloudera’s ecosystem of software included in CDH4.1 such as Cloudera Manager, Hadoop, Hue, Beeswax, Oozie, Pig, Hive, Impala, and HBase to process hundreds of Terabytes of information.

·         Responsible for overseeing the batch data workflow processes created using Spring Batch, Hive QL against Hive and Impala, Java against HBase, and direct Shell commands; coordinate efforts with Operations and Infrastructure to verify that the cluster nodes are configured and performing in accordance to user expectations.

·         Oversaw the troubleshooting and escalation of any issues during the installation and configuration process; consulted and assisted in researching hardware setup related issues; and forwarded all remaining blocking problems to the experts both in online groups, or directly to Cloudera.

·         Facilitated a proof of concept for an open source data integration solution from Talend and evaluated alternate solutions such as Python-based Celery and Java-based Spring Batch.

·         Participated in and documented the details of the data ingestion system from the high level framework to the technical implementation and consequently helped design the inclusion of reporting workflows and machine learning behavior algorithms in a modularized fashion.

·         Coordinated vendor meetings and presentations between the team and Talend for their Enterprise Data Integration Suite, Cloudera for their Cloudera Manager Tool, and Radoop for their Data Analytics, Mining, and Machine Learning tool.

·         Pushed for an initiative to move to real-time computational analysis using Storm in addition to Kafka as data messaging bus.

 

Odesus, Los Angeles, CA              06/2012 – 07/2012

Client: Fox Digital Media, Playa Vista, CA

Big Data Architect              (contract)

 

·         Investigated data extraction methods and solutions to funnel all the data exports from the multitude of third party analytics vendors into a centralized data repository. It was decided to use bash with some Python, cron with Jenkins to schedule, coordinate, and monitor the extractions.

·         Due to the nature of the project being as a POC plus budgetary constraints, Amazon AWS was commissioned as an IaaS for S3 storage, Hadoop/Hbase/Hive for analytics on EMR, and MySQL for the dashboard on RDS.

·         Overviewed the data workflow processes contained in the Java MapReduce, Hive QL, and Python code.

·         Google Analytics were seen as a more capable, cost effective solution, but legal and political concerns halted any further trials.

·         Other trial efforts using Google’s services were Cloud Storage as a data repository and BigQuery for analytics.

·         Touched on converting, changing the current SDLC methodology from agile to kanban.

 

Level Studios, El Segundo, CA              02/2012 – 06/2012

Client: Activision, Santa Monica, CA

Data Solutions Architect              (contract)

 

·         Analyzed, designed, and implemented an automated method of data extraction for reports, ad campaigns, and analytics to measure effectiveness.

·         The data sources involved were a combination of Hadoop, Oracle, and Infobright.

·         Initial investigation into using the Talend DI tool to simplify and automate the process was implemented but was ultimately abandoned.

·         Bash scripts and cron jobs were sufficient for their needs.

 

Accenture, El Segundo, CA              2011 – 2012

Data Platform Architect

 

·         Planned, designed, and oversaw the implementation of the big data infrastructure using open source technologies such as Hadoop, Flume, Sqoop, Pig, Hive, and HBase to process petabytes of information.

·         Led the facilitating a proof of concept ETL solution to populate HDFS using the open source data integration tool from Pentaho.

·         Oversaw the troubleshooting and escalation of any issues in Hadoop, HBase, Hive, and Flume by triaging Map/Reduce or Flume Source/Sink coding tickets; consulted and assisted in researching configuration related issues; and forwarding all remaining blocking problems to the experts both within the company, online, or to Cloudera.