Initially, I led Data Engineering on the Data Platform team where I spearheaded our architecture and best practices around Data Engineering and ML Training Infrastructure. As Knock grew and hired more engineers I transitioned into leadership. My focus today is on leading Infrastructure, Data, ML, and Strategic platform initiatives within all of Engineering.- Upgraded all our CI/CD pipelines to use terraform for deployment, and created a reusable CI package to streamline and standardize all our pipelines in CircleCI.- Extended an internal workflow service to support AWS EMR for spark data jobs and Tensorflow ML training data pipelines. I later led a migration project to Apache Airflow on EKS (Kubernetes) for both CPU and GPU data workloads.

- Led the company to become data driven, and led the build out of a centralized data warehouse in snowflake with DBT/Fivetran/Airflow/EMR/Tableau/Spark.Designed and built systems for ingesting millions of real estate listings and listing photos from 35 MLSs (Multiple Listing Service) using Golang, SQS, S3, DynamoDB, and Kinesis.- Designed and led a team to normalize the MLS data using Spark streaming and ingest it into downstream systems of ElasticSearch, Snowflake, and Athena/s3 including a rest service that is our internal data search engine.

- Analyzed internal and public gis data/tabular data to develop an algorithm to define internal markets for sharding ML model inference and training.- Designed and built our own internal geocoder using Apache Spark, Openaddresses, lib postal and postgis. I later led a project to port it to DynamoDB for better scaling/performance under a high load with additional cost savings.

Lead the company in Architectural design choices for new and old systems. Find solutions to problems in old code bases while driving the a new vision and direction for the new code bases. The company is small so when not working on architecture, I joined the team in development and operations sprint work.Accomplishments- Helped migrate code off of parse.com platform to the open source parse-server framework.- Containerized all parts for the system for easy deployment and versioning of deployments.- Created a repeatable deployment process so the system can auto scale based on load, and system health.- Instrumented code with application metrics to find and remove bottlenecks in code and to monitor overall system health.- Encapsulated system dependences to enable testing/staging environments. - Lead the team to understand and implement how to make unit testable code with dependency injection and interfaces, and making sure that all new code has unit tests. - Lead the team to understand the importance of when making changes to old code to see if there is a way to make it unit testable though small code refactoring.Environment: Golang, Java, Scala, Javascript, Python, Android, Git, Maven, Gradle, Ubuntu 14.04/16.04, Docker, AWS, Opsworks, Chef, Fabric, MongoDB, Redis, InfluxDB, DCOS, and Apache Spark.

RoleDesign and Build world class high-volume, real-time data ingestion and processing systems and advanced analytics on big data platforms. Research, develop, optimize, and innovate frameworks and patterns for enterprise scale data analysis and computations as part of our big data and Internet of Things initiatives. Lead the implementation of Hadoop platform strategy development by creating architecture blueprints, validating designs and providing recommendations on the enterprise platform strategic roadmap. Presented technology solutions to senior leadership and influenced architectures that will lead the transformation of the IoT data analytics platform and IT overall.Accomplishments -Researched, Designed, Lead, and Built an IoT project to create a system to compete with Cisco System’s Location Services (MSE). Brought the cost of entry to Location Analytics down from 50k to under 1k. Used off the shelf low cost router hardware with custom build of OpenWrt and Apache Storm to do real-time wifi triangulation. -Researched, Designed, Lead, Built a Big Data lake used to generate adhoc reports for business and customers based on IoT location data. This system has reduced cost 11.5 times over the -previous system, and can handle 50 times the load. System used Apache Kafka, and Apache Spark running on an Hortonworks Data Platform cluster. -Designed and Built a monitoring and alert system used to monitor current system load and alert of outage events and forecast outages and Load. -Performed system administration and DevOps, deployed Hadoop in production along with other open source systems on linux vms running in Windows Azure IaaS.Environment: Scala, Java, Python, Apache Spark, Apache Kafka, Apache Storm, Linux, Git, Maven, Windows Server 2012, Windows Azure, Android, iOS, Azure Storage, Azure Service Bus, MVC3/4/5, web API 1/2, OWIN, Visual Studio 2010/2012/2013, SVN, Tfs, Node.js, Shell, and C#

- Worked on a web application for managing the billing system supported and developed new features. - Worked in a day-to-day support roll in billing helping bill USC customers. - Created scripts in Perl and Shell for automating customer change of address process with USPS. - Often used Perl for testing work and other day-to-day tasks. - Reverse engineered a transaction system to sync 2 different billing systems, solving 15000 billing errors per month. - Worked on a project to move a piece of the billing system off a Mainframe onto Unix and Cognos systems.Environment: Hpux/Linux, Oracle, Java, Struts 2, Hibernate, PL/SQL, Perl, Shell, BIRT, Cognos, and ClearCase

- Designed a new company topic idea system for storing and viewing ideas across the business. - Created and normalized a database for topic ideas. - Designed the look and feel of the application.Environment: ASP.NET, IIS, VB, and SQL Server

- Built web software for and designed the User Interface controls. - Designed and implemented the user admin pages. - Created data-driven web pages using ASP.NET.Environment: ASP.NET, IIS, C#, DotNetNuke and SQL Server

Give 1.5 hour tours which included 2-mile dog sled ride, a tour of dog sled kennel, review all the equipment, and puppies!!! When I was not give tours or dog sled rides I was taking care of the 180 dogs and puppies.