Automated and productionalized Halo mobile app engagement machine learning models with CI/CD pipelines using CDK, Lambda, SageMaker, ECR, CodeDeploy, CodeCommit, S3, and Redshift.Built RESTful API's using CDK, Lambda, API Gateway, Route53, S3, and Redshift for Halo mobile analytics reporting platform.Managed and maintained 300TB Redshift cluster for the Alexa Music team with near real-time Kinesis Firehose data stream sources originating from Alexa devices.Designed and built python-based ETL dependency framework to enable integration of multiple internal Amazon data systems. Framework was utilized to enable thousands of ETL jobs over 5+ years.Built serverless ETL using CDK (Typescript), Step Functions, Glue, Lambda functions, and S3.Work with Python, Redshift, Aurora Postgres, AWS CDK, Typescript, S3, Kinesis Firehose, Lambda, Step Functions, internal Amazon systems (Apollo/Brazil/DJS/Datanet/EDX/Cradle/Andes/Carnaval).Collaborated with Economists + Compensation Consultants and built data warehouse & pipelines to measure results of wage change experiment across Amazon fulfillment centers.

Created data pipeline and ranking algorithm for Last Call feature using python, Redshift, DynamoDB, S3, EC2, & Hive. The idea is to leverage web traffic data + real estate operational data to create a bidding war on Redfin listings. Blog Post: https://www.redfin.com/blog/2015/02/redfin-last-call-service-informs-you-when-to-make-an-offer-soo.htmlLed development on compensation system responsible for calculating payment of over $50M in 2015. Article: http://www.inman.com/2014/11/15/redfin-has-risen-still-an-industry-disrupter/Modeled, designed and developed enterprise data warehouse.Created MLS data integration using Java to launch Redfin in Park City, Utah market.Led data modeling and ETL migration moving from GoodData to Tableau as an enterprise BI tool. Integrated Workday HR data into data warehouse by developing a Java SOAP API client.Used Pentaho DI, git, bash, Tableau, node, SQL Server, Postgres, JIRA

Led Business Intelligence Development for a Fortune 500 company.Worked with vendors and led Data Virtualization software acquisition.Developed dashboards and reports using complex SQL, Java, and JavaScript.Managed operation of Solaris server clusters running Tomcat with Apache and fronted by Big IP load balancers.Designed and developed an Enterprise Data Warehouse using transactional systems, Informatica PowerCenter, bash scripts, and DB2. The project used a rapid development methodology and had daily scrums with users.Collaborated with cross-functional business and IT groups to define specifications and develop logical and physical data models for BI applications.Provided ongoing technical assistance and guidance to developers located at worldwide company branches.Led HADR project for an enterprise BI web-app with 2000+ users. Created governance utilities to diminish impact of poorly designed code.Integrated Google Maps API with enterprise reporting web-app using jQuery, Mozilla Rhino, and DB2.Implemented Java single-sign-on solution to integrate BI web app with enterprise portal.Migrated legacy Data Warehouse from SQL Server to DB2 and ETL from DTS to Informatica.Saved users and development team 20+ hours per month by developing an automated CRM data scrubbing process using Informatica ETL and bash scripts. Regularly used Solaris, Eclipse, Firebug, vim, cron, ssh, JConsole, MS Project, and trello.