Kolin Ohi Email & Phone Number

New York City, New York, Us

Figure Technologies creates innovative consumer financial solutions for home improvement, debt consolidation and retirement. Its mission is to build and promote innovative financial products on blockchain that benefit consumers and eliminate the rent-seeking, illiquidity and other inefficiencies in current financial markets.As Senior Director of Data Engineering, I am responsible for the data architecture, engineering, and operations that support analytics, reporting, and data science. As one of the earliest members of the data team, I built our data infrastructure from the ground up, including:- Created streaming data ingest from event-driven Kafka architecture, with on-the-fly process that translates protocol buffer schemas defined in multiple microservice applications into BigQuery data structures that evolve in real time. - Designed and built Python ETL of operational message formats into dimensional data lake, with schemas optimized for reporting and analytics.- Implemented Airflow on Google Kubernetes Engine, enabling automation of many thousand jobs each day.- Built ETL framework in Python, pyspark, and BigQuery, providing the reporting benefits of a dimensional data model (slowly changing dimensions, accumulating snapshots, and fact tables) without the overhead and expense of a data warehouse or ETL tools.- Automated data ingest from credit agencies and property data brokers, wiring legacy batch flat file formats and technologies (sftp, etc.) into our architecture and allowing data scientists to focus on direct mail models instead of file logistics and data hygiene.

Austin, Texas, Us

Created a unified, cost efficient data platform to power restaurant operations and analytics. Ingesting and processing data from multiple legacy applications using Spark, Spark Streaming, Kinesis, and EMR.- Drove a data lake strategy to consolidate data assets from HotSchedules’ multiple companies and technology platforms. Enabled a centralized view of data to drive new product features and insights for restaurant customers.- Created and executing on a roadmap to transition data from expensive or legacy platforms (Redshift, SQLServer, MongoDB) to more scalable and cost-efficient platforms (Presto/Spark on EMR and centralized s3 storage).Accomplishments included:- Built stream ingestion for data from multiple restaurant points of sale data and its transformation into a common schema for centralized analytics.- Built forecasting engine on Spark to enable granular sales forecasting efficiently at scale.

San Francisco, California, Us

Responsible for team that rebuilt Eventbrite’s data architecture. Achievements included:- Optimizing costs and performance by transitioning from reserved CDH5 instances running hive ETL to EMR clusters running pyspark transformations. Split workloads between transient and persistent clusters with shared storage on s3.- Rebuilding entire infrastructure from the bottom up using terraform and ansible. Environment can now be deployed seamlessly in any Availability Zone.- Redesigning analytic schemas with a hybrid data lake/ dimensional approach, enabling ‘single source of truth’ without slowing down new product engineering.- Rolling out new job scheduling on Luigi, SLA and operations monitoring on Elasticsearch/Logstash/Kibana, and Python code framework for data pipelines.- Refactoring data ingestion pipelines (Salesforce, MySQL, Google Analytics, Kafka) from improvised homegrown solutions to maintainable code that follows industry best practices.- Enabling data science team to explore data in spark and python with Zeppelin; enabling them to deploy models to production with spark code framework.- Designing schema repository/metadata approach- Establishing operational processes; publishing and reporting on SLAs and operational incidents

Lead multiple teams implementing Hotwire's Business Intelligence initiatives in US, Eastern Europe, and India. Scope includes ETL and database/hadoop development (Informatica, Oracle Exadata, hive/pig/java), front-end (MicroStrategy & Qlikview), analytic tools (R Studio Server, hue, etc.), ETL operations, release management, and QA.

Led BI architecture and analytic database, ETL, and front-end development. Responsibilities included:• Drove design and tool selection: led selection and implementation of Oracle Exadata, Control-M, Hadoop, Qlikview, and MicroStrategy Visual Insights.• Selected and deployed Hadoop solution (CDH4) to augment existing environment, delaying a multi-million dollar database expansion. • Expanded Hotwire’s analytic capabilities with RHadoop and other tools.• Managed and mentored a team of engineers and architects, both locally and offshore.• Deals Engine: owner of projects that automated and optimized Hotwire merchandizing, including deals displays on the website, deals promoted in email marketing, and feeds sent to marketing partners.• Worked with pricing and data science teams to implement statistical algorithms for optimized displays, sorting, and pricing.• Provided analytic support to all areas of the company; handled all analyses that required large data volumes or advanced SQL or PL/SQL.

Roles included: Principal, Analytics; Manager, Product Analytics; Senior Analyst.

Detroit, Michigan, Us

San Francisco, California, Us

Pleasanton, Ca, Us

Ba, Summa Cum Laude, English Literature, Comparative Literature

Ma, English